<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Monika Singh on Medium]]></title>
        <description><![CDATA[Stories by Monika Singh on Medium]]></description>
        <link>https://medium.com/@Monika_Singh?source=rss-376aecb22a0d------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*W3vkRSbKz-Vf64nfHSqxQw.jpeg</url>
            <title>Stories by Monika Singh on Medium</title>
            <link>https://medium.com/@Monika_Singh?source=rss-376aecb22a0d------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 13:58:35 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@Monika_Singh/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Role of IITs in Shaping the Future of Indian Students]]></title>
            <link>https://medium.com/@Monika_Singh/role-of-iits-in-shaping-the-future-of-indian-students-1149a23bf79e?source=rss-376aecb22a0d------2</link>
            <guid isPermaLink="false">https://medium.com/p/1149a23bf79e</guid>
            <category><![CDATA[i̇i̇t]]></category>
            <category><![CDATA[education]]></category>
            <dc:creator><![CDATA[Monika Singh]]></dc:creator>
            <pubDate>Fri, 30 Jun 2023 22:19:17 GMT</pubDate>
            <atom:updated>2023-06-30T22:19:17.782Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*c48zC43WBVP5dlgJ.jpg" /></figure><p>The Indian Institutes of Technology (IITs) play a crucial role in shaping the future of Indian college students in numerous ways. These top-of-the-line establishments have a robust recognition for academic excellence, research, and innovation, and they have been instrumental in generating some of the brightest minds in India. Here are a few key roles that IITs play:</p><ol><li><strong>Quality Education: </strong>IITs are renowned for their rigorous instructional applications. They provide undergraduate, postgraduate, and doctoral programs in diverse engineering, technology, and sciences fields. The curriculum is designed to offer a stable basis in theoretical information and practical skills, preparing college students for expert success.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/371/0*zmK1YzwWSNclQ8E9" /></figure><p><strong>2. Research and Innovation: </strong>IITs are at India&#39;s leading edge in scientific research and technological improvements. They behavior modern-day studies in numerous regions which include nanotechnology, biotechnology, synthetic intelligence, robotics, and renewable energy, amongst others. This research culture encourages college students to discover new ideas, make contributions to know-how creation, and broaden progressive answers to real-global problems.<br>Here are some notable examples of research and innovation from IITs:</p><h4>IIT Madras and Shakti Processors:</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/655/1*ECiyfufl0JQAUgGHxO_NSw.png" /></figure><p>These processors are based on the RISC-V structure and feature the capability to energy an extensive variety of programs, along with cell devices, IoT devices, and supercomputers. This improvement showcases IIT Madras’ know-how in semiconductor layout and its dedication to advancing the indigenous era.</p><h4>IIT Kharagpur and Rural Healthcare Solutions:</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/768/1*_KiCfNElF3i-zWiCURgC3A.jpeg" /></figure><p>IIT Kharagpur has undertaken several research projects to increase healthcare solutions for rural regions. Projects consist of low-price clinical devices, telemedicine systems, and cell healthcare programs to enhance healthcare get admission to and transport in faraway regions. These innovations mirror IIT Kharagpur’s emphasis on the use of technology to cope with healthcare disparities.</p><h4>IIT Kanpur and COVIRAP:</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/700/1*6BehbsbOI4A7EIX1o_xazg.jpeg" /></figure><p>IIT Kanpur developed COVIRAP, a low-price, portable, and fast diagnostic era for COVID-19. This innovation makes use of a paper-strip-primarily based totally diagnostic approach, imparting brief and correct results. COVIRAP has the capability to enhance COVID-19 testing capabilities, in particular in resource-restrained settings. This improvement verified IIT Kanpur’s dedication to leveraging the era for addressing socially demanding situations.</p><p>These examples highlight the various research and innovation projects undertaken by IITs, starting from advanced technology improvement to answers addressing societal needs.</p><p><strong>3. Entrepreneurship and Start-up Ecosystem: </strong>IITs offer assets and support through incubation centers, entrepreneurship cells, and mentorship programs, encouraging students to transform their thoughts into feasible start-ups. Many hit Indian start-ups, which include Flipkart, Ola, and Paytm, had been based with the aid of using IIT alumni.</p><blockquote><strong>Satya Nadella (IIT Delhi) and Skype:</strong> Satya Nadella, the CEO of Microsoft, studied at IIT Delhi earlier than pursuing better training abroad. While at Microsoft, he performed a key function in the acquisition of Skype.</blockquote><figure><img alt="Satya Nadella, the CEO of Microsoft," src="https://cdn-images-1.medium.com/max/770/1*ttbWtbL81lAbybccLwNd2w.jpeg" /></figure><blockquote><strong>Binny Bansal and Sachin Bansal (IIT Delhi) and Flipkart:</strong> Binny Bansal and Sachin Bansal, the co-founders of Flipkart, started their entrepreneurial adventure at the same time as analyzing at IIT Delhi.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/900/1*ndsceR2OHeEMT3DjQBTZdw.jpeg" /></figure><blockquote><strong>Deepinder Goyal and Pankaj Chaddah (IIT Delhi) and Zomato: </strong>Deepinder Goyal and Pankaj Chaddah, each IIT Delhi alumni, co-founded Zomato, an internet meals shipping and eating place discovery platform. Zomato has become one of India’s most successful tech start-ups, remodeling the food enterprise.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*AbvOrVGZfoi8lAFmNjkbsQ.jpeg" /></figure><blockquote><strong>Bhavish Aggarwal and Ankit Bhati (IIT Bombay) and Ola: </strong>Bhavish Aggarwal and Ankit Bhati, each graduate from IIT Bombay, co-based Ola, India’s leading ride-hailing service. Convenient platform for booking rides through a mobile app, transforming the way people commute in Indian cities.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/750/1*GvEAmIAgsbwjAXMXBVnPBQ.jpeg" /></figure><p><strong>4. Industry Collaboration:</strong> IITs preserve strong ties with industries and collaborate on research projects, internships, and placements. This collaboration allows enterprise exposure, sensible training, and internships for students, improving their employability. Many reputed countrywide and global organizations actively recruit from IITs, presenting appealing activity possibilities and professional growth.</p><p>Here are some examples of industry collaborations involving IITs:</p><ol><li>Tata Motors and IIT Bombay</li><li>Microsoft Research and IIT Delhi</li><li>Boeing and IIT Kanpur</li><li>General Electric (GE) and IIT Madras</li><li>IBM and IIT Delhi</li><li>Mahindra &amp; Mahindra and IIT Gandhinagar</li></ol><p><strong>5. Global Recognition: </strong>IITs revel in global popularity for academic excellence and feature continuously ranked as some of the pinnacle engineering establishments worldwide. This popularity draws global college students, fostering cultural range and presenting Indian students with exposure to a global learning environment. Moreover, IIT alumni have made large contributions globally, strengthening India’s image as a hub of technical expertise.</p><blockquote><strong>Sundar Pichai (IIT Kharagpur):</strong> Sundar Pichai, the CEO of Google and one of the most influential tech executives globally, is an alumnus of IIT Kharagpur. His success story serves as a notion for limitless Indian students, showcasing the ability for accomplishing fantastic heights withinside the technology industry.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1qjaL0OLYDYHYYvqnDK04A.jpeg" /></figure><blockquote><strong>N. R. Narayana Murthy (IIT Kanpur):</strong> N. R. Narayana Murthy, the co-founder of Infosys, graduated from IIT Kanpur. He appeared as a pioneer of the Indian IT enterprise and has performed an essential function in remodeling India’s software program services sector. Murthy’s journey from IIT to entrepreneurship has prompted several students to pursue their entrepreneurial aspirations.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/929/1*eIUE1WCG2bbrZM0ftGF2IA.jpeg" /></figure><p><strong>6. Social Impact:</strong> IITs actively make a contribution to the social and monetary improvement of India. They adopt numerous outreach programs, community initiatives, and social entrepreneurship initiatives to deal with societal challenges. Through studies, technological innovations, and collaborations, IITs attempt to find sustainable answers to urgent troubles in regions together with healthcare, education, sanitation, and renewable energy.</p><blockquote><strong>Smart Village Project by IIT Delhi: </strong>The Smart Village Project showcased IIT Delhi’s commitment to utilizing technology and expertise for the betterment of society. By collaborating with local stakeholders and implementing sustainable solutions, the project had a direct and positive impact on the lives of villagers, empowering them and enhancing their quality of life.</blockquote><h4>Conclusion:</h4><p><strong>Overall, IITs have performed a transformative role in shaping the future of Indian students by providing them with exceptional education, fostering studies and innovation, selling entrepreneurship, facilitating enterprise collaboration, gaining worldwide recognition, and creating an effective social impact. These establishments maintain to encourage and empower students to excel, make a contribution to society, and power India’s increase withinside the fields of science, engineering, and technology.</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1149a23bf79e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Title: The Importance of Extracurricular Activities in Boarding Schools in Pune]]></title>
            <link>https://medium.com/@Monika_Singh/title-the-importance-of-extracurricular-activities-in-boarding-schools-in-pune-7a1edd7fff24?source=rss-376aecb22a0d------2</link>
            <guid isPermaLink="false">https://medium.com/p/7a1edd7fff24</guid>
            <category><![CDATA[evaluation]]></category>
            <category><![CDATA[boarding-school]]></category>
            <category><![CDATA[boarding-schools-in-pune]]></category>
            <category><![CDATA[pune]]></category>
            <category><![CDATA[education]]></category>
            <dc:creator><![CDATA[Monika Singh]]></dc:creator>
            <pubDate>Sat, 17 Jun 2023 07:16:27 GMT</pubDate>
            <atom:updated>2023-06-17T07:16:27.249Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/748/0*QmIrAkSa2rZcb_75.jpg" /><figcaption>image-1</figcaption></figure><h3><strong>Introduction:</strong></h3><p>Boarding colleges in Pune provide a completely unique academic revel that is going past academics. Alongside rigorous instructional programs, those establishments area a robust emphasis on extracurricular sports. These sports, starting from sports activities to arts and network providers, play a crucial function withinside the holistic improvement of college students.</p><h3><strong>Boarding Schools in Pune</strong></h3><p>Boarding colleges in Pune provide a completely unique academic enjoy that is going past conventional day colleges. With a focal point on holistic improvement, those establishments offer nurturing surroundings for college kids to thrive academically, socially, and emotionally.</p><p><strong>1. Academic Excellence:</strong><br>Boarding colleges in Pune are famous for their dedication to educational excellence. They provide rigorous curricula, committed college, and well-geared-up lecture rooms to offer college students a sturdy basis of knowledge. The consciousness of lecturers prepares college students for destiny educational hobbies and equips them with important thinking, problem-solving, and study abilities.</p><p><strong>2. Residential Life and Character Development:</strong><br>The residential factor of boarding colleges performs an extensive function in shaping college students’ person and fostering a non-public boom. Living far from domestic teaches college students independence, self-reliance, and adaptability. Boarding colleges in Pune make sure of secure and supportive surroundings, fostering an experience of network and inspiring college students to embody variety and broaden lifelong friendships.</p><p><strong>3. Holistic Approach to Education:</strong><br>Boarding colleges in Pune emphasize a holistic method of education, spotting the significance of nurturing college students’ physical, mental, and emotional well-being. They offer an extensive variety of extracurricular activities, which include sports, arts, music, and clubs, which permit college students to discover their interests, broaden talents, and domesticate lifestyle abilities consisting of teamwork, management, and time management.</p><p><strong>4. Cultural and Global Exposure:</strong><br>Boarding colleges in Pune regularly have various pupil bodies, permitting college students to engage with friends from unique cultures and backgrounds. This publicity fosters an appreciation for variety, complements cultural sensitivity, and promotes worldwide awareness. Schools regularly arrange cultural events, worldwide change programs, and collaborative projects, enriching college students’ knowledge of the world.</p><p><strong>5. Dedicated Faculty and Support Systems:</strong><br>Boarding colleges in Pune rent certified and committed college contributors who offer individualized interest and aid to college students. The close-knit network permits significant teacher-pupil relationships, mentorship, and guidance. Additionally, colleges provide complete aid systems, which include counseling services, educational aid, and pastoral care, making sure the well-being and achievement of college students.</p><p><strong>6. Focus on Character Values and Leadership:</strong><br>Boarding colleges in Pune location a sturdy emphasis on personal improvement and management abilities. They foster an experience of responsibility, integrity, and moral values among college students. Leadership opportunities, consisting of pupil councils, clubs, and network carrier initiatives, permit college students to broaden management qualities, social responsibility, and an experience of civic duty.</p><h3>The Importance of Extracurricular Activities in Boarding Schools in Pune</h3><p>The evaluative evaluation highlights the strengths of boarding colleges in Pune and their extensive contributions to the holistic improvement of college students, getting ready to come to be well-rounded people able to deal with destiny-demanding situations with confidence. In this blog, we can discover the significance of extracurricular sports in boarding colleges in Pune and the way they make contributions to shaping well-rounded individuals.</p><p><strong>1. Enhancing Personal Growth:</strong></p><p>Extracurricular sports offer possibilities for college kids to discover their interests, find out new talents, and increase their critical existence talents. Engaging in sports which include sports activities, music, drama, and debate fosters self-confidence, teamwork, leadership, and time control talents. These stories make contributions to non-public increases and assist college students to end up well-rounded individuals.</p><p>The blog efficiently highlights the numerous methods wherein extracurricular sports facilitate private growth, emphasizing the improvement of important existence capabilities that amplify past the classroom.</p><p><strong>2. Promoting Physical Fitness and Health:</strong></p><p>Regular participation in bodily sports is critical for retaining proper health. Boarding colleges in Pune apprehend the importance of bodily health and offer a huge variety of sports activities and outside sports. From soccer and basketball to swimming and yoga, those sports sell a lively lifestyle, bodily well-being, and usual health amongst college students.</p><p>The blog efficiently highlights the importance of bodily health and the position of extracurricular sports in selling a wholesome lifestyle. It emphasizes the huge variety of alternatives to be had in boarding faculties in Pune to cater to students’ numerous interests.</p><p><strong>3. Nurturing Creativity and Artistic Expression:</strong></p><p>Extracurricular sports in arts, which include painting, music, dance, and theater, offer avenues for college kids to specific their creativity. These sports inspire self-expression, imagination, and innovation. They permit college students to discover their creative talents, increase essential wondering talents, and foster a deep appreciation for the arts.</p><p>The blog efficaciously highlights the position of extracurricular sports in nurturing creativity and creative expression. It emphasizes the advantages of undertaking inventive endeavors, consisting of more advantageous creativeness and important wondering abilities.</p><p><strong>4. Building Social Skills and Relationships:</strong></p><p>Participating in extracurricular sports gives college students the possibility to interact and collaborate with their peers. Whether it is via group sports activities or membership sports, college students learn how to paint together, speak effectively, and construct significant relationships. These social talents are worthwhile in growing robust interpersonal relationships and getting ready college students for destiny endeavors.</p><p>The blog efficiently emphasizes the social thing of extracurricular sports and the way they make a contribution to the improvement of vital social skills. It highlights the collaborative nature of those sports and their effective effect on constructing relationships.</p><p><strong>5. Developing a Sense of Responsibility and Civic Engagement:</strong></p><p>Boarding colleges in Pune regularly emphasize network providers and social duty. Engaging in volunteer paintings, fundraisers, and network initiatives instills an experience of empathy, compassion, and duty toward others. Students research the significance of giving lower back to society, growing a robust experience of civic engagement, and contributing definitely to their communities.</p><p>The blog efficiently highlights the importance of extracurricular sports in instilling an experience of obligation and civic engagement. It recognizes the boarding schools’ dedication to a network carrier and the nice effect it has on students’ man or woman development.</p><p><strong>6. Balancing Academic Pressure:</strong></p><p>Extracurricular sports offer a much-wished stability to the instructional strain that scholars might also additionally revel in boarding colleges. By conducting sports they enjoy, college students can lessen stress, decorate intellectual well-being, and preserve a wholesome paintings-existence stability. Participation in extracurricular sports has additionally been related to stepped-forward instructional overall performance and expanded motivation.</p><p>The blog successfully addresses the problem of educational stress and highlights how extracurricular sports make a contribution to preserving a healthy balance. It helps its claims with applicable studies findings.</p><h3>Conclusion:</h3><p>Extracurricular sports in boarding colleges in Pune play a pivotal function in shaping well-rounded individuals. They provide college students with possibilities for non-public increase, bodily health, creativity, social interaction, and civic engagement. These sports now no longer most effectively supplement the instructional curriculum however additionally make contributions to the holistic improvement of college students.</p><p>By encouraging participation in extracurricular sports, boarding colleges in Pune foster supportive surroundings wherein college students can discover their passions, increase critical talents, and domesticate a lifelong love for learning.</p><p>image-1 <a href="https://boardingschoolsofindia.com/wp-content/uploads/2017/12/4cc3abd5bf.jpg">https://boardingschoolsofindia.com/wp-content/uploads/2017/12/4cc3abd5bf.jpg</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7a1edd7fff24" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Fraudulent Transaction Case Study]]></title>
            <link>https://medium.com/@Monika_Singh/fraudulent-transaction-case-study-e1922d395fdc?source=rss-376aecb22a0d------2</link>
            <guid isPermaLink="false">https://medium.com/p/e1922d395fdc</guid>
            <category><![CDATA[roc-curve]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[supervised-learning]]></category>
            <dc:creator><![CDATA[Monika Singh]]></dc:creator>
            <pubDate>Tue, 13 Sep 2022 04:26:38 GMT</pubDate>
            <atom:updated>2022-09-13T04:26:38.749Z</atom:updated>
            <content:encoded><![CDATA[<h3>Fraud Detection Case Study</h3><h4>1. Importing Data and Libraries</h4><p>Understanding the data</p><p>step — maps a unit of time in the real world. In this case 1 step is 1 hour of time. Total steps 744 (30 days simulation).</p><p>type — CASH-IN, CASH-OUT, DEBIT, PAYMENT and TRANSFER.</p><p>amount — amount of the transaction in local currency.</p><p>nameOrig — customer who started the transaction</p><p>oldbalanceOrg — initial balance before the transaction</p><p>newbalanceOrig — new balance after the transaction</p><p>nameDest — customer who is the recipient of the transaction</p><p>oldbalanceDest — initial balance recipient before the transaction. Note that there is not information for customers that start with M (Merchants).</p><p>newbalanceDest — new balance recipient after the transaction. Note that there is not information for customers that start with M (Merchants).</p><p>isFraud — This is the transactions made by the fraudulent agents inside the simulation. In this specific dataset the fraudulent behavior of the agents aims to profit by taking control or customers accounts and try to empty the funds by transferring to another account and then cashing out of the system.</p><p>isFlaggedFraud — The business model aims to control massive transfers from one account to another and flags illegal attempts. An illegal attempt in this dataset is an attempt to transfer more than 200.000 in a single transaction.</p><pre># Suppressing Warnings<br>import warnings<br>warnings.filterwarnings(&#39;ignore&#39;)</pre><pre># Importing Pandas and NumPy<br>import pandas as pd, numpy as np</pre><pre># Importing the dataset<br>Fraud_df = pd.read_csv(&quot;Fraud.csv&quot;)<br>Fraud_df.head()</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*uHBmhgYOskbIh3ev8Q_-eQ.png" /></figure><pre># check the dimensions of the dataframe<br>Fraud_df.shape</pre><pre>(6362620, 11)</pre><pre># see the type of each column<br>Fraud_df.info()</pre><pre>&lt;class &#39;pandas.core.frame.DataFrame&#39;&gt;<br>RangeIndex: 6362620 entries, 0 to 6362619<br>Data columns (total 11 columns):<br> #   Column          Dtype  <br>---  ------          -----  <br> 0   step            int64  <br> 1   type            object <br> 2   amount          float64<br> 3   nameOrig        object <br> 4   oldbalanceOrg   float64<br> 5   newbalanceOrig  float64<br> 6   nameDest        object <br> 7   oldbalanceDest  float64<br> 8   newbalanceDest  float64<br> 9   isFraud         int64  <br> 10  isFlaggedFraud  int64  <br>dtypes: float64(5), int64(3), object(3)<br>memory usage: 534.0+ MB</pre><pre># look at the statistical aspects of the dataframe<br>Fraud_df.describe(include = &quot;all&quot;)</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*vHSyTNnkrkLU59RkiAHdOw.png" /></figure><h4>2. Data Cleaning and Preperation</h4><pre>Fraud_df.columns</pre><pre>Index([&#39;amount&#39;, &#39;oldbalanceOrg&#39;, &#39;newbalanceOrig&#39;, &#39;oldbalanceDest&#39;,<br>       &#39;newbalanceDest&#39;, &#39;isFraud&#39;, &#39;isFlaggedFraud&#39;, &#39;type_CASH_OUT&#39;,<br>       &#39;type_DEBIT&#39;, &#39;type_PAYMENT&#39;, &#39;type_TRANSFER&#39;],<br>      dtype=&#39;object&#39;)</pre><h4>Checking for duplicate values</h4><pre>duplicatevalues = Fraud_df.duplicated()<br>Fraud_df[duplicatevalues]</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*YhpDDsO5HdKPiNAIS55TNA.png" /></figure><pre># Checking the duplicate values<br>Fraud_df.duplicated().sum()</pre><pre>0</pre><h4>Checking for Null values</h4><pre>Fraud_df.isna().sum()</pre><pre>step              0<br>type              0<br>amount            0<br>nameOrig          0<br>oldbalanceOrg     0<br>newbalanceOrig    0<br>nameDest          0<br>oldbalanceDest    0<br>newbalanceDest    0<br>isFraud           0<br>isFlaggedFraud    0<br>dtype: int64</pre><h4>Checking for outliers</h4><pre># Checking for outliers in the continuous variables<br>num_df = Fraud_df[[&#39;amount&#39;,&#39;oldbalanceOrg&#39;,&#39;newbalanceOrig&#39;,&#39;oldbalanceDest&#39;, &#39;newbalanceDest&#39;]]</pre><pre># Checking outliers at 25%, 50%, 75%, 90%, 95% and 99%<br>num_df.describe(percentiles=[.25, .5, .75, .90, .95, .99])</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*3B2izZeSnz03l6F1YAlaQw.png" /></figure><pre>plt.figure(figsize=(15,7))<br>Fraud_df[[&#39;amount&#39;,&#39;oldbalanceOrg&#39;,&#39;newbalanceOrig&#39;,&#39;oldbalanceDest&#39;, &#39;newbalanceDest&#39;]].boxplot()</pre><pre>&lt;AxesSubplot:&gt;</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*SHH1rMIZkp2B_7QsC0HPTA.png" /></figure><pre>def remove_outlier(col):<br>    sorted(col)<br>    Q1,Q3=np.percentile(col,[25,75])<br>    IQR=Q3-Q1<br>    lower_range= Q1-(1.5 * IQR)<br>    upper_range= Q3+(1.5 * IQR)<br>    return lower_range, upper_range</pre><pre>for feature in Fraud_df[[&#39;amount&#39;,&#39;oldbalanceOrg&#39;,&#39;newbalanceOrig&#39;,&#39;oldbalanceDest&#39;, &#39;newbalanceDest&#39;]]: <br>    lr,ur=remove_outlier(Fraud_df[feature])<br>    Fraud_df[feature]=np.where(Fraud_df[feature]&gt;ur,ur,Fraud_df[feature])<br>    Fraud_df[feature]=np.where(Fraud_df[feature]&lt;lr,lr,Fraud_df[feature])</pre><pre>plt.figure(figsize=(15,7))<br>Fraud_df[[&#39;amount&#39;,&#39;oldbalanceOrg&#39;,&#39;newbalanceOrig&#39;,&#39;oldbalanceDest&#39;, &#39;newbalanceDest&#39;]].boxplot()</pre><pre>&lt;AxesSubplot:&gt;</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*DBYrhG507XDQtINhru1B1A.png" /></figure><p>Dropping “nameOrig”, “nameDest” columns as Name columns cannot be a significant contributer for Fraud Detection.</p><pre>Fraud_df.drop(columns=[&quot;nameOrig&quot;,&quot;nameDest&quot;],inplace=True)</pre><pre>Fraud_df.head()</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*ckcSuqQyqhv7Fu1If5sxzA.png" /></figure><pre>Fraud_df[&quot;type&quot;].value_counts()</pre><pre>CASH_OUT    2237500<br>PAYMENT     2151495<br>CASH_IN     1399284<br>TRANSFER     532909<br>DEBIT         41432<br>Name: type, dtype: int64</pre><pre>sns.countplot(Fraud_df[&quot;type&quot;]);</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*lSPQamLwSjgUrri8JkOvKg.png" /></figure><pre># Histogram of Amount feature based on fraud<br>plt.figure(figsize=(8,6))<br>sns.histplot(data=Fraud_df[Fraud_df[&#39;isFraud&#39;]==1],x=&#39;amount&#39;,kde=True)<br>plt.title(&#39;Histogram of Amount of Fradulent Transactions&#39;)<br>plt.show()</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*86oS24gNb3UCjhaJzGNMpw.png" /></figure><p>Distribution is positively skewed</p><h4>Creating Dummy Variables</h4><pre>temp=pd.get_dummies(Fraud_df[&quot;type&quot;],drop_first=True,prefix=&quot;type&quot;)<br>temp</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*RremWQzDQDmc63nVhU28Qw.png" /></figure><pre>Fraud_df = pd.concat([Fraud_df,temp],axis=1)<br>Fraud_df</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*yy9_MbhQHogd6ysw4spccg.png" /></figure><pre>Fraud_df.drop(columns=[&quot;type&quot;,&quot;step&quot;],inplace=True)<br>Fraud_df</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*__mWTolEt3fziCFfzu_gAQ.png" /></figure><h4>Checking Correlation</h4><pre># Importing matplotlib and seaborn<br>import matplotlib.pyplot as plt<br>import seaborn as sns<br>%matplotlib inline</pre><pre>Fraud_df.corr()</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*5wlYHQ9gZhpC3vJekjI64w.png" /></figure><pre># Let&#39;s see the correlation matrix <br>plt.figure(figsize = (20,10))        # Size of the figure<br>sns.heatmap(Fraud_df.corr(),annot = True)<br>plt.show()</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*SHp67h8nbBMhBz2A731STw.png" /></figure><p>None of the columns have a very good correlation with the variable isFraud</p><h4>3. Feature Scaling</h4><pre>from sklearn.preprocessing import MinMaxScaler</pre><pre>y = Fraud_df[&quot;isFraud&quot;]<br>X = Fraud_df.drop(columns=[&quot;isFraud&quot;])</pre><pre>scaler=MinMaxScaler()<br>X_scaled=scaler.fit_transform(X)</pre><h4>4. Test-Train Split</h4><pre>from sklearn.model_selection import train_test_split</pre><pre>X_train, X_test, y_train, y_test = train_test_split(X,y,train_size=0.7,test_size=0.3,random_state=50)</pre><h4>5. Model Building</h4><h4>Logistic regression model</h4><pre>from sklearn.linear_model import LogisticRegression</pre><pre>reg =LogisticRegression()<br>reg.fit(X_train,y_train)<br>y_pred=reg.predict(X_test)</pre><h4>Performance Metrics</h4><pre>from sklearn.metrics import confusion_matrix,accuracy_score,classification_report,f1_score,recall_score,precision_score</pre><pre>confusion = confusion_matrix(y_test,y_pred)<br>confusion</pre><pre>array([[1902306,    4016],<br>       [    532,    1932]], dtype=int64)</pre><pre>print(&quot;Accuracy:&quot;,accuracy_score(y_test, y_pred))<br>print(&quot;Precision:&quot;,precision_score(y_test, y_pred))<br>print(&quot;Recall:&quot;,recall_score(y_test, y_pred))</pre><pre>Accuracy: 0.9976173337398745<br>Precision: 0.32481506388702086<br>Recall: 0.7840909090909091</pre><pre>report =classification_report(y_test,y_pred)<br>print(report)</pre><pre>precision    recall  f1-score   support<br><br>           0       1.00      1.00      1.00   1906322<br>           1       0.32      0.78      0.46      2464<br><br>    accuracy                           1.00   1908786<br>   macro avg       0.66      0.89      0.73   1908786<br>weighted avg       1.00      1.00      1.00   1908786</pre><pre>def plot_conf_mat(y_test, y_pred):<br>  fig, ax = plt.subplots(figsize=(10,7))<br>  ax = sns.heatmap(confusion_matrix(y_test, y_pred), annot=True, cbar=False)<br>  plt.xlabel(&quot;True label&quot;)<br>  plt.ylabel(&quot;Predicted label&quot;)<br><br>plot_conf_mat(y_test, y_pred)</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*KdEZTmvWPJp-aoxvE0ajQQ.png" /></figure><h4>Metrics beyond simply accuracy</h4><pre>TP = confusion[1,1] # true positive <br>TN = confusion[0,0] # true negatives<br>FP = confusion[0,1] # false positives<br>FN = confusion[1,0] # false negatives</pre><pre># Let&#39;s see the sensitivity of our logistic regression model<br>TP / float(TP+FN)</pre><pre>0.7840909090909091</pre><pre># Let us calculate specificity<br>TN / float(TN+FP)</pre><pre>0.9978933254717723</pre><pre># Calculate false postive rate - predicting churn when customer does not have churned<br>print(FP/ float(TN+FP))</pre><pre>0.002106674528227655</pre><pre># positive predictive value <br>print (TP / float(TP+FP))</pre><pre>0.32481506388702086</pre><pre># Negative predictive value<br>print (TN / float(TN+ FN))</pre><pre>0.9997204176078047</pre><h4>6. Plotting the ROC Curve</h4><p>An ROC curve demonstrates several things:</p><p>It shows the tradeoff between sensitivity and specificity (any increase in sensitivity will be accompanied by a decrease in specificity). The closer the curve follows the left-hand border and then the top border of the ROC space, the more accurate the test. The closer the curve comes to the 45-degree diagonal of the ROC space, the less accurate the test.</p><pre>y_prob=reg.predict_proba(X_test)[:,1]<br>y_prob</pre><pre>array([1.96363649e-132, 1.01782222e-117, 0.00000000e+000, ...,<br>       2.01572858e-095, 3.43577104e-006, 3.43901538e-008])</pre><pre>def draw_roc( actual, probs ):<br>    fpr, tpr, thresholds = metrics.roc_curve( actual, probs,<br>                                              drop_intermediate = False )<br>    auc_score = metrics.roc_auc_score( actual, probs )<br>    plt.figure(figsize=(5, 5))<br>    plt.plot( fpr, tpr, label=&#39;ROC curve (area = %0.2f)&#39; % auc_score )<br>    plt.plot([0, 1], [0, 1], &#39;k--&#39;)<br>    plt.xlim([0.0, 1.0])<br>    plt.ylim([0.0, 1.05])<br>    plt.xlabel(&#39;False Positive Rate or [1 - True Negative Rate]&#39;)<br>    plt.ylabel(&#39;True Positive Rate&#39;)<br>    plt.title(&#39;Receiver operating characteristic example&#39;)<br>    plt.legend(loc=&quot;lower right&quot;)<br>    plt.show()<br><br>    return None</pre><pre>from sklearn import metrics<br>fpr, tpr, thresholds = metrics.roc_curve( y_test, y_prob, drop_intermediate = False )</pre><pre>draw_roc(y_test, y_prob)</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*utAKjQpWc_Km7A2doIE-Mw.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e1922d395fdc" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Forest Fire Prediction for NASA MODIS Satellite Data]]></title>
            <link>https://medium.com/@Monika_Singh/forest-fire-prediction-for-nasa-modis-satellite-data-3cf9fe273d94?source=rss-376aecb22a0d------2</link>
            <guid isPermaLink="false">https://medium.com/p/3cf9fe273d94</guid>
            <category><![CDATA[statistics]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[regression]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[mls]]></category>
            <dc:creator><![CDATA[Monika Singh]]></dc:creator>
            <pubDate>Sun, 21 Aug 2022 13:04:57 GMT</pubDate>
            <atom:updated>2022-08-21T13:04:57.202Z</atom:updated>
            <content:encoded><![CDATA[<h3>Forest Fire Prediction</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yjWHhJomv2PZii94YtOY_A.jpeg" /><figcaption>Forest Fire in India</figcaption></figure><h3><strong>Introduction</strong></h3><p>Natural resources play an essential role in living, but natural hazards such as tropical cyclones, earthquakes, tsunamis, landslides, and drought in different areas hinder the sustainable development of humans and living things. An increase in global warming increases the probability of climate-related disasters like heat waves, drought, and wildfire.</p><p>Wildfire is an unpredictable, unplanned, and uncontrolled fire that can be caused by various human activities such as coal seams, campfire(bonfire), smoking, Debris and Arson burning, unauthorized burning, firework, shooting, exploding targets or various natural phenomena such as lightning, meteor, volcanic activities, high temperature that can burn the natural vegetation like the forest, savannah, and other natural ecosystem. In this study forest fire information dataset from NASA is taken for the experiment</p><h4>Data Source</h4><p>The data source used in this study is obtained from NASA’s Fire Information for Resource Management System (FIRMS) satellite data and observations of wildfire cases in India 2021 from NASA’s MODIS (Moderate Resolution Imaging Spectroradiometer) instrument.</p><h4>Reading and Understanding the Data</h4><p>Import NumPy and Pandas and read the MODIS dataset of forest fire</p><pre># Supress Warnings<br><br>import warnings<br>warnings.filterwarnings(&#39;ignore&#39;)</pre><pre>import numpy as np<br>import pandas as pd</pre><pre>df = pd.read_csv(&quot;modis_2021_India.csv&quot;)</pre><pre># Check the head of the dataset<br>df.head()</pre><p><strong>Dataset Features and Description</strong></p><ol><li>latitude</li><li>longitude</li><li>brightness = brightness tempertaure of fire in pixel measured in Kelvin</li><li>scan = scan pixel size</li><li>track = track pixel size</li><li>acq_date = acquisition date</li><li>acq_time = acquisition time</li><li>satellite = Acua and Terra satellite as A and T</li><li>instrument =MODIS</li><li>confidence = Collection of intermediate algorithm quantities used in the detection process and estimate range between 0% -100%</li><li>version = Version identifies the collection and data processing source</li><li>bright_t31 = brightness temperature of fire from Channel 31 that is measured in Kelvin</li><li>frp =Fire Radiative Power in megawatts unit</li><li>daynight = D and N as Day and Night</li><li>type = 0 as presumed vegetation, 1 as active volcano , 2 as static land source 3 as detection for offshore</li></ol><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*Fl0GBeZVCvacvoIAwLOaxg.png" /></figure><pre>df.shape</pre><pre>(111267, 15)</pre><pre>df.info()</pre><pre>&lt;class &#39;pandas.core.frame.DataFrame&#39;&gt;<br>RangeIndex: 111267 entries, 0 to 111266<br>Data columns (total 15 columns):<br> #   Column      Non-Null Count   Dtype  <br>---  ------      --------------   -----  <br> 0   latitude    111267 non-null  float64<br> 1   longitude   111267 non-null  float64<br> 2   brightness  111267 non-null  float64<br> 3   scan        111267 non-null  float64<br> 4   track       111267 non-null  float64<br> 5   acq_date    111267 non-null  object <br> 6   acq_time    111267 non-null  int64  <br> 7   satellite   111267 non-null  object <br> 8   instrument  111267 non-null  object <br> 9   confidence  111267 non-null  int64  <br> 10  version     111267 non-null  float64<br> 11  bright_t31  111267 non-null  float64<br> 12  frp         111267 non-null  float64<br> 13  daynight    111267 non-null  object <br> 14  type        111267 non-null  int64  <br>dtypes: float64(8), int64(3), object(4)<br>memory usage: 12.7+ MB</pre><pre>df.columns</pre><pre>Index([&#39;latitude&#39;, &#39;longitude&#39;, &#39;brightness&#39;, &#39;scan&#39;, &#39;track&#39;, &#39;acq_date&#39;,<br>       &#39;acq_time&#39;, &#39;satellite&#39;, &#39;instrument&#39;, &#39;confidence&#39;, &#39;version&#39;,<br>       &#39;bright_t31&#39;, &#39;frp&#39;, &#39;daynight&#39;, &#39;type&#39;],<br>      dtype=&#39;object&#39;)</pre><pre>df.isnull().sum()</pre><pre>latitude      0<br>longitude     0<br>brightness    0<br>scan          0<br>track         0<br>acq_date      0<br>acq_time      0<br>satellite     0<br>instrument    0<br>confidence    0<br>version       0<br>bright_t31    0<br>frp           0<br>daynight      0<br>type          0<br>dtype: int64</pre><pre>df.describe()</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*mhgypoNNo0xEu89Mny9HvQ.png" /></figure><h3>Visualizing the Data</h3><p>Visualize our data using matplotlib and seaborn</p><pre>import matplotlib.pyplot as plt<br>import seaborn as sns</pre><ol><li><strong>Visualizing Numeric Variables</strong></li></ol><p>Let’s make a pairplot of all the numeric variables</p><p>Seaborn is more integrated for working with Pandas data frames. It extends the Matplotlib library for creating beautiful graphics with Python using a more straightforward set of methods.The show() function simply displays the result of a series of operations. Because of that, we can gradually fine-tune a lot of details in the figure.</p><pre>sns.pairplot(df)<br>plt.show()</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*R9axDhbBt6AT01NVBLI6jg.png" /></figure><p><strong>2. Visualizing Categorical Variables</strong></p><p>As you might have noticed, there are a few categorical variables as well. Let’s make a boxplot for some of these variables.</p><pre>plt.figure(figsize=(20, 12))<br>plt.subplot(2,2,1)<br>sns.boxplot(x = &#39;satellite&#39;, y = &#39;confidence&#39;, data = df)<br>plt.subplot(2,2,2)<br>sns.boxplot(x = &#39;daynight&#39;, y = &#39;confidence&#39;, data = df)</pre><pre>&lt;AxesSubplot:xlabel=&#39;daynight&#39;, ylabel=&#39;confidence&#39;&gt;</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*h7yyr4EutHhaNmQdeDDVjA.png" /></figure><p>Let’s check the correlation coefficients to see which variables are highly correlated</p><pre>plt.figure(figsize=(10, 10))<br>sns.heatmap(df.corr(),annot=True,cmap=&#39;viridis&#39;,linewidths=.5)</pre><pre>&lt;AxesSubplot:&gt;</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*Ezg7ZuuLKoFLI-TEJ6hyhw.png" /></figure><pre>sns.barplot(x=&#39;acq_date&#39;,y=&#39;frp&#39;,data=df)</pre><pre>&lt;AxesSubplot:xlabel=&#39;acq_date&#39;, ylabel=&#39;frp&#39;&gt;</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*xtiopZe48B8yZCxKAL2q6w.png" /></figure><pre>df_topaffected=df.sort_values(by=&#39;frp&#39;,ascending=False)<br>df_topaffected.head(10)</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*qgIq4rR7nAkbW_4IuFsw5g.png" /></figure><h3>Data Preparation</h3><p>Data can be in the form of text, videos, images, or more which may contain various missing values, null values, and unnecessary data. This unstructured data may interrupt while training the model or will through some errors. So, before training, the unstructured data must be formatted or cleaned to implement the Machine Learning algorithms effectively</p><p>Data cleaning is filling the missing values, cleaning the irrelevant data, and transforming data using normalization and feature selection.</p><pre>df = df.drop([&#39;track&#39;], axis = 1)<br><br>df = df.drop([&#39;instrument&#39;, &#39;version&#39;], axis = 1)</pre><pre>df[&#39;satellite&#39;] = df[&#39;satellite&#39;].map({&#39;Terra&#39;:0,&#39;Aqua&#39;:1})<br>df[&#39;daynight&#39;] = df[&#39;daynight&#39;].map({&#39;D&#39;:0,&#39;N&#39;:1})<br>df[&#39;month&#39;] = df[&#39;acq_date&#39;].apply(lambda x:int(x.split(&#39;-&#39;)[1]))<br>df = df.sample(frac=0.2)<br>df = df.reset_index().drop(&quot;index&quot;, axis=1)</pre><pre>df.head()</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*6_QJhePTU7fKp5epvwsn8A.png" /></figure><pre>df.shape</pre><pre>(22253, 13)</pre><pre>y = df[&#39;confidence&#39;]<br>firedf = df.drop([&#39;confidence&#39;, &#39;acq_date&#39;], axis = 1)</pre><pre>plt.figure(figsize=(10, 10))<br>sns.heatmap(firedf.corr(),annot=True,cmap=&#39;viridis&#39;,linewidths=.5)</pre><pre>&lt;AxesSubplot:&gt;</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*GlWnJI5cOJeRfwfklfd_Xg.png" /></figure><pre>firedf.head()</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*EE6X0ikHtABJfZMgHA_THQ.png" /></figure><h3>Splitting the Data into Training and Testing Sets</h3><p>Splitting the dataset into two part of training set (i.e. 70% of data) and testing set(i.e. 30% of data)</p><pre>X = df[[&#39;latitude&#39;,&#39;longitude&#39;,&#39;month&#39;,&#39;brightness&#39;,&#39;scan&#39;,&#39;acq_time&#39;,&#39;bright_t31&#39;,&#39;daynight&#39;]]<br>y = df[&#39;frp&#39;]<br>from sklearn.model_selection import train_test_split<br>X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3)</pre><h4>Evaluation Matrix</h4><p>To measures model performance using the training set of the Machine Learning model, some of the loss functions used for the evaluation, such as MSE and MAE among the actual values and the predicted values, can be calculated by importing the scikit-learn python library as it consists of the predefines functions like mean_squared_error() function and mean_absolute_error() fuction. It also provides the predefined process for R squared score as r2_score().</p><ol><li><strong>Mean Squared Error (MSE)</strong></li></ol><p>MSE can be calculated mathematically from the average squared difference of the actual valus from the dataset and predicted values</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/265/1*gkOVTzchOQcyqkQJzYEVrA.png" /></figure><p>y =actual y value</p><p>ŷ =predicted y value</p><p>n = no. of observations</p><p><strong>2. Mean Absolute Error (MAE)</strong></p><p>Mean Absolute Error can be mathematically calculated as the average difference of the value that is predicted from the data and the actual values</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/260/1*oONyzoxyVOt7nx-uMwdDdw.png" /></figure><p>yi =actual y value</p><p>ŷi =predicted y value</p><p>n = no. of observations</p><p><strong>3. Root Mean Square Error (RMSE)</strong></p><p>Mathematically RMSE can be calculated as the square root of the average squared difference of the values predicted and the actual value from the data, or we can say it is the square root value of MSE.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/265/1*MVSgOfPDiXkMNtC-0eyatw.png" /></figure><p>yi =actual y value</p><p>ŷi =predicted y value</p><p>n = no. of observations</p><p><strong>4. R Squared Score</strong></p><p>R squared score can be calculated from the equation given below:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/398/1*hNNRH8lX3GY53icqe5sgyA.png" /></figure><p>yi =actual y value</p><p>ŷi =predicted y value</p><p>ӯ = mean of y value</p><p>N = no. of observations</p><h4>Algorithm Used</h4><h4>1. Gradient Boosting Regressor</h4><p>Gradient boosting Regression calculates the difference between the current prediction and the known correct target value. To train the model, import GradientBoostingRegressor class. We can use the function fit() with which we can fit the values for Xtrain and values for ytrain.</p><pre>from sklearn.ensemble import GradientBoostingRegressor<br>from sklearn.metrics import mean_absolute_error as mae<br>from sklearn.metrics import mean_squared_error as mse<br>from sklearn.metrics import r2_score<br>model1 = GradientBoostingRegressor(n_estimators = 100, learning_rate=0.1,<br>                                  max_depth = 10, random_state = 0, loss = &#39;ls&#39;)<br>model1.fit(X_train, y_train)<br>y_pred = model1.predict(X_test)<br>print (&#39;MSE =&#39;,mse(y_pred, y_test))<br>print (&#39;RSME =&#39;,np.sqrt(mse(y_pred, y_test)))<br>print (&#39;MAE =&#39;,mae(y_pred, y_test))<br>print (&#39;R2_score =&#39;,r2_score(y_pred, y_test))<br><br>print(&quot;Performance ofGBR Model R^2 metric {:.5f}&quot;.format(model1.score(X_train,y_train)))<br>print(&quot;GBR Accuracy, {:.5f}%&quot;.format(model1.score(X_test,y_test)*100))</pre><pre>MSE = 449.63478569897046<br>RSME = 21.20459350468597<br>MAE = 3.2932809652939343<br>R2_score = 0.9227647882671393<br>Performance ofGBR Model R^2 metric 0.99982<br>GBR Accuracy, 92.86853%</pre><h4>2. Decision Tree Regressor</h4><p>Another approach for supervised learning is the Decision Tree Algorithm for solving decision-related problems,which can be used for both Regression and Classification. It consists of various nodes in the form of branches, such as Root node(just like roots of the tree), Interior node, and leaf node(just like leafs of the tree), as this algorithm is a tree structure-based algorithm.</p><p>To train the model, import DecisionTreeRegressor class. We can use the function fit() with which we can fit the values for Xtrain and values for ytrain.</p><pre>from sklearn.tree import DecisionTreeRegressor as dtr<br>reg = dtr(random_state = 42)<br>reg.fit(X_train,y_train)<br>Y_pred = reg.predict(X_test)<br>print(&quot;MSE = &quot;,mse(Y_pred, y_test))<br>print (&#39;RSME =&#39;,np.sqrt(mse(Y_pred, y_test)))<br>print(&quot;MAE =&quot;,mae(Y_pred,y_test))<br>print(&quot;R2 score =&quot;,r2_score(Y_pred,y_test))<br>print(&quot;Performance of Decision Tree Regressor Model R^2 metric {:.5f}&quot;.format(reg.score(X_train,y_train)))<br>print(&quot;Decision Tree Regressor Accuracy, {:.5f}%&quot;.format(reg.score(X_test,y_test)*100))</pre><pre>MSE =  362.66616536848414<br>RSME = 19.043795981066488<br>MAE = 4.693229478729778<br>R2 score = 0.9316181035003857<br>Performance of Decision Tree Regressor Model R^2 metric 1.00000<br>Decision Tree Regressor Accuracy, 94.24790%</pre><p>Graphical representation of the efficiency of both the ML Algorithm used to build the predictive model for Forest Fire Prediction is compared with the help of Radar Chart A radar chart is a graphical method used for displaying multivariate data in the form of a two-dimensional chart of three or more quantitative variables represented on axes starting from the same point. (Source: Wikipedia)</p><pre>categories = [&#39;Training score&#39;, &#39;Testing score&#39;, &#39;RMSE&#39;, &#39;MAE&#39;,&#39;r2&#39;,&#39;_&#39;]<br><br><br><br>Decision_Tree_Regressor = [100,94.24, 19.4,4.69, 0.93,0]<br><br><br>Gradient_Boosting_Regressor = [99.97, 92.86, 21.20,3.29, 0.92,0]<br><br>label_loc = np.linspace(start=0, stop=2 * np.pi, num=len(Decision_Tree_Regressor))<br><br>plt.figure(figsize=(8, 8))<br>plt.subplot(polar=True)<br><br>plt.plot(label_loc, Decision_Tree_Regressor , label=&#39;Decision Tree Regression&#39;)<br><br><br>plt.plot(label_loc, Gradient_Boosting_Regressor, label=&#39;Gradient Boosting Regressor&#39;)<br><br>plt.title(&#39;Algorithms comparison for MODIS dataset&#39;, size=20)<br>lines, labels = plt.thetagrids(np.degrees(label_loc), labels=categories)<br><br>plt.legend()<br>plt.show()</pre><figure><img alt="png" src="https://cdn-images-1.medium.com/proxy/1*2fD57gSOiP89LyAKw9wiNg.png" /></figure><p><strong>Conclusion</strong></p><p>Our motive is to estimate the susceptibility and the risk of fire occurrence due to natural factors or human activities such as road development &amp; construction and human-forest interfaces.</p><p>The proposed approach has achieved high prediction accuracy with ML algorithms like Gradient Boosting Regressor i.e. 93% and Decision tree Regressor i.e. 94% for NASA FIRMS dataset. In the future, considering the NASA FIRMS, operational data used for the study and the proposed method can create a real-time learning environment to predict forest fire risk effectively.</p><h3><strong>Profile</strong></h3><p><a href="https://www.linkedin.com/in/monika-singh-b022951b8/">https://www.linkedin.com/in/monika-singh-b022951b8/</a></p><p>Click the link below to visit the code discussed above :</p><p><a href="https://github.com/MonikaSingh105/ForestFire_Prediction">GitHub - MonikaSingh105/ForestFire_Prediction: Supervised learning based prediction model for prediction of forest fire in India using NASA FIRMS active data of forest fire in India 2021 from satellite MODIS.</a></p><p><strong>Dataset</strong></p><p><a href="https://firms.modaps.eosdis.nasa.gov/download/">NASA-FIRMS</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3cf9fe273d94" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>