<?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 Tanush Siva on Medium]]></title>
        <description><![CDATA[Stories by Tanush Siva on Medium]]></description>
        <link>https://medium.com/@tanush.siva?source=rss-e341653a8265------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*BWLxuhPRnHiDbXla</url>
            <title>Stories by Tanush Siva on Medium</title>
            <link>https://medium.com/@tanush.siva?source=rss-e341653a8265------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 08:26:00 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@tanush.siva/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[Quantum Annealing and QUBO for Smart Water Distribution]]></title>
            <link>https://medium.com/@tanush.siva/quantum-annealing-and-qubo-for-smart-water-distribution-3dc900fcaf64?source=rss-e341653a8265------2</link>
            <guid isPermaLink="false">https://medium.com/p/3dc900fcaf64</guid>
            <category><![CDATA[quantum-computing]]></category>
            <category><![CDATA[optimization-algorithms]]></category>
            <category><![CDATA[quantum-annealing]]></category>
            <category><![CDATA[water-distribution]]></category>
            <category><![CDATA[quantum-machine-learning]]></category>
            <dc:creator><![CDATA[Tanush Siva]]></dc:creator>
            <pubDate>Sun, 27 Apr 2025 02:40:33 GMT</pubDate>
            <atom:updated>2025-04-27T02:40:33.320Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*6ERZQdYK88Gwn7p8.jpg" /><figcaption>Image by Gerd Altmann from Pixabay</figcaption></figure><p>Most real world optimization problems are defined by a complex solution landscape full of local minima, causing classical approaches to stumble. True optimization has remained out of reach…until now.</p><p>Using the power of quantum tunneling and parallelism, quantum annealing is now navigating solution spaces that were once thought impossible to traverse.</p><p>Droughts are spreading across the world, water resources are shrinking, and many traditional water allocation methods are struggling to keep up. In regions where every drop of water counts, ineffective water distribution will lead to economic loss and humanitarian crisis. Classical optimization models take impractical amounts of time to find solutions and can settle for suboptimal approaches. As time ticks down and water becomes scarce, faster allocation strategies are no longer optional. Quantum annealing, paired with a QUBO model, will unlock faster and smarter solutions for water distribution.</p><h3><strong>What is Quantum Annealing?</strong></h3><p>Picture this: You’ve been asked to find the deepest valley in a mountain range. But there is one problem: you’re blindfolded. If you were to follow a classical approach to this problem, you would go step by step downhill, continuing to move lower. Eventually, you could get stuck in a smaller dip, and misrecognize it as the deepest point without seeing the entire landscape. Using a quantum approach, on the other hand, allows you to tunnel through these mountains and head straight towards the deepest valley.</p><p>Quantum annealing is a quantum computing technique designed to solve optimization problems (like water allocation). By using quantum fluctuations, this process is able to find the global minimum of an objective function over a set of candidate solutions.</p><h4><strong>Qubit Evolution During Annealing</strong></h4><p>Initially, each qubit in the system starts in superposition, representing the states of both 0 and 1. As the quantum annealing process runs, an energy barrier is gradually raised, transforming the solution landscape into a <strong>double well potential</strong>. Applying an external magnetic field will <strong>tilt </strong>the double well, adjusting the probability of the qubit being in the 0 or 1 state. Tilting the well increases the probability of the qubit settling in a valley with the lower energy minimum, helping the system reach the most optimal solution for the optimization problem.</p><h4><strong>Coupling Between Qubits</strong></h4><p>Quantum annealing can also allow the “linking” of qubits through a <strong>coupler</strong>. These can create interactions between qubits to make them end up in the same, or opposite classical states. Via programmatically tuned qubit biases and coupling, the qubits in the system eventually become entangled to mirror the structure of the optimization problem. By the end of the annealing process, the qubits will be in a state which represents the global energy minimum of the optimization problem (the most optimal solution).</p><h3><strong>What is a QUBO Model?</strong></h3><p>As described in the last article, a QUBO (Quadratic Unconstrained Binary Optimization) model is a way to represent and solve a combinatorial optimization problem. The goal is to find an optimal solution from a discrete set of possibilities. The model is made of an <strong>objective function</strong> as well as a set of <strong>constraints</strong>.</p><ol><li>Objective Function: Defines the goal of the optimization problem. The model aims to minimize the value of this function, which represents a performance metric for the given solution.</li><li>Constraints: Enforce the “rules” a given solution should follow. Represented as penalty terms which are added to the objective function, increasing its value when certain constraints are violated.</li></ol><h3><strong>How Do QUBO and Annealing Work Together?</strong></h3><p>Once a problem is in the form of a QUBO model, it must be mapped into an equivalent <strong>Ising </strong>model, a type of <strong>Hamiltonian</strong> which describes the interaction between a system of binary variables (representing spins). In the Ising model, each qubit represents a spin variable, linear terms correspond to the qubit biases (local magnetic field strengths), and quadratic terms define the coupling strengths between specific pairs of qubits.</p><p>During the annealing process, an initial Hamiltonian which has an easily preparable ground state is evolved to a final Hamiltonian whose ground states encode the solution to the optimization problem. At the end of the process, the qubits settle into a position which minimizes the energy of the system, representing a near-optimal solution.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*LwTFFY4IZXRLwdZk.jpg" /><figcaption>Image by Jose Antonio Alba from Pixabay</figcaption></figure><p>With almost two-thirds of the world’s population facing severe water scarcity for at least a month every year, the stakes have never been higher. Many classical approaches are too slow and limited for the scale of present challenges. With the rise of quantum computing, quantum annealing will offer a way to defend our resources and find smarter solutions when every drop of water counts.</p><ul><li><a href="https://docs.dwavequantum.com/en/latest/quantum_research/quantum_annealing_intro.html">What is Quantum Annealing? - Python documentation</a></li><li><a href="https://docs.dwavequantum.com/en/latest/quantum_research/qubo_ising.html?utm_source=chatgpt.com">QUBOs and Ising Models - Python documentation</a></li><li><a href="https://www.unicef.org/wash/water-scarcity">Water scarcity</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3dc900fcaf64" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Using QUBO For Equitable Water Allocation]]></title>
            <link>https://medium.com/@tanush.siva/using-qubo-for-equitable-water-allocation-473a2838a3bd?source=rss-e341653a8265------2</link>
            <guid isPermaLink="false">https://medium.com/p/473a2838a3bd</guid>
            <category><![CDATA[drought]]></category>
            <category><![CDATA[quantum-algorithms]]></category>
            <category><![CDATA[quantum-computer]]></category>
            <category><![CDATA[quantum-computing]]></category>
            <category><![CDATA[quantum-machine-learning]]></category>
            <dc:creator><![CDATA[Tanush Siva]]></dc:creator>
            <pubDate>Sat, 12 Apr 2025 20:16:48 GMT</pubDate>
            <atom:updated>2025-04-27T03:24:22.623Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*6KsrT1lWc4mQrzx0.jpg" /><figcaption>Image by Brent Olson from Pixabay</figcaption></figure><p>As drought continues to worsen and water supplies dwindle, nations around the world have begun to face challenges with <strong>food security </strong>and <strong>human health</strong>. The people of <strong>Sindh</strong>, <strong>Pakistan </strong>depend on the Indus River for their water supply. With the onset of drought and climate change, <strong>fair </strong>and <strong>equitable </strong>water allocation has become a serious issue for this region. If this problem is left unaddressed, this could ultimately lead to water wars amongst the people of this nation and many others around the world. Using the power of <strong>quantum computing </strong>and <strong>QUBO </strong>models, we can ensure the equitable distribution of water among drought-affected regions, preventing water wars and other resource conflicts. (<a href="https://www.dawn.com/news/1880389"><strong>https://www.dawn.com/news/1880389</strong></a>)</p><h3><strong>What Are QUBO Models?</strong></h3><p>A QUBO (<strong>Q</strong>uadratic <strong>U</strong>nconstrained <strong>B</strong>inary <strong>O</strong>ptimization) model is a way to represent and solve a<strong> combinatorial optimization</strong> problem. The goal is to find an optimal solution from a discrete set of possibilities. A good way to picture this situation is imagining a <strong>puzzle</strong>. Each piece in this puzzle represents a part of a potential solution to the optimization problem, and the goal is to fit these pieces together to form the best possible outcome.</p><p>The goal of the QUBO model is to minimize a <strong>quadratic objective function</strong>.</p><ol><li><strong>Quadratic</strong>: QUBO models can use linear or pairwise quadratic terms. The mathematics of the optimization problem itself will determine if quadratic terms are needed, or if linear terms are enough to solve the problem.</li><li><strong>Unconstrained</strong>: QUBO models will penalize solutions which violate constraints but will not insist on a solution to satisfy all constraint terms.</li><li><strong>Binary</strong>: The variables used in the QUBO model are of binary nature, meaning each variable can take one of two values: 0 or 1.</li><li><strong>Optimization</strong>: These models are rooted in optimization mathematics, which finds the best answer for a problem based on a set of criteria which conflict with each other. QUBO models excel in combinatorial optimization problems and are used in a variety of applications in this field.</li></ol><h4><strong>What Advantages Do They Offer Over Classical Optimization Models?</strong></h4><p>Pairing a QUBO model with a quantum optimization technique like <strong>quantum annealing</strong> or <strong>QAOA</strong> (Quantum Approximate Optimization Algorithm) offers several advantages over classical optimization models.</p><ol><li><strong>Complex Solution Landscapes: </strong>When solving combinatorial optimization problems, classical computers often struggle with maintaining a diverse solution set. Specifically, in discrete high-dimensional search spaces with many local minima, classical algorithms often get stuck, and are unable to find the global minimum (representing the best solution set). Quantum optimization techniques overcome this limitation by escaping local minima via <strong>quantum tunneling.</strong></li><li><strong>Speed: </strong>Classical algorithms also have issues with efficiency when solving optimization problems. In large-scale optimization problems, where the number of potential solutions grow significantly, classical methods require more time to explore the solution space. Quantum optimization techniques offer an exponential speedup over classical algorithms by leveraging quantum superposition and entanglement. This parallelism allows quantum computers to evaluate many solution sets simultaneously.</li></ol><h4><strong>The Objective Function</strong></h4><p>In a QUBO model, the objective function is what defines the goal of the optimization problem. The model aims to minimize the value of this function, which represents a performance metric for the given solution. Effectively, the model wants to find the solution with the “least cost”. The function consists of linear and quadratic terms, and is represented as a sum of weighted binary variables. The weight of the objective function is typically tuned depending on the type of optimization problem at hand.</p><h4><strong>The Constraint Terms</strong></h4><p>The constraint terms of a model enforce the “rules” a given solution should follow. They are represented as penalty terms which are added to the objective function, increasing its value when certain constraints are violated. The strength of each constraint is controlled by a tunable weight term.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*smaNUOQjTc3dvGbG.jpg" /><figcaption>Image by <a href="https://pixabay.com/users/thedigitalartist-202249/?utm_source=link-attribution&amp;utm_medium=referral&amp;utm_campaign=image&amp;utm_content=3679893">Pete Linforth</a> from <a href="https://pixabay.com//?utm_source=link-attribution&amp;utm_medium=referral&amp;utm_campaign=image&amp;utm_content=3679893">Pixabay</a></figcaption></figure><p>Stay tuned for part two of this article, covering quantum optimization techniques such as quantum annealing and QAOA!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=473a2838a3bd" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Nystrom Quantum Kernel Estimation And Drought]]></title>
            <link>https://medium.com/@tanush.siva/nystrom-quantum-kernel-estimation-and-drought-d4bc8b40194d?source=rss-e341653a8265------2</link>
            <guid isPermaLink="false">https://medium.com/p/d4bc8b40194d</guid>
            <category><![CDATA[quantum-computer]]></category>
            <category><![CDATA[quantum-computing]]></category>
            <category><![CDATA[computer-science]]></category>
            <category><![CDATA[high-school]]></category>
            <category><![CDATA[drought]]></category>
            <dc:creator><![CDATA[Tanush Siva]]></dc:creator>
            <pubDate>Tue, 11 Feb 2025 22:17:05 GMT</pubDate>
            <atom:updated>2025-02-11T22:17:05.429Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*ISs4pkAEbmKT3CxW.jpg" /><figcaption>Image by svklimkin from Pixabay</figcaption></figure><p>Droughts don’t just suddenly happen–they slowly build up, silently creeping across landscapes until it’s too late. When drought hits a region, its impact on a community is devastating, causing crops to wither away and damaging human health.</p><p>Predicting these disasters beforehand is critical. Traditional machine learning models (which use classical computing) often struggle to predict drought in a timely manner and are prone to data overfitting.</p><p>How to solve these problems, you ask? Enter the world of quantum computing. Leveraging the power of <strong>Nyström Quantum Kernel Estimation </strong>in combination with classical random forests will allow us to more efficiently and accurately predict droughts.</p><h3><strong>Recap: What Are Random Forests and Decision Trees?</strong></h3><p>As described in the last article, decision trees are commonly used hierarchical tree-like structures in machine learning to make decisions, predictions, and classifications given a dataset. They are made up of root, internal, and leaf nodes, as well as branches.</p><p>Random forests take a random sample of the given data, and create an ongoing set of decision trees. These models help reduce the data overfitting commonly induced by classical decision trees. The random feature selection ensures the same data is not used for every tree, making the model less sensitive and reducing the correlation between the trees.</p><h4><strong>What Are the Issues with Random Forests?</strong></h4><p>The crux of classical random forest models comes with the poor performance of split nodes within the decision trees. The hyperparameters of the split function need to be manually selected or algorithmically tuned. For a split function to be effective, it must amplify the separation of classes down the tree, but still generalize, as repeated splitting can lead to data overfitting.</p><h3><strong>What is Nyström Quantum Kernel Estimation?</strong></h3><p>In simple terms, Nyström Quantum Kernel Estimation<strong> </strong>(NQKE) is a method that speeds up calculations in machine learning models by selecting a small set of key points, called pivot points, instead of comparing every data point. It combines the Nyström approximation method with quantum kernels to estimate large kernel matrices. NQKE can be used to improve the split function of random forests. By employing the split function in the form of a support vector machine that employs a quantum kernel, we are able to generate a hyperplane in a higher dimensional quantum feature space. This allows the classes in the data set to be readily distinguishable.</p><p>My Quantum project aims to fight against drought, as it is now threatening much of the world. There are others too, trying in their own special way to do the same. Researchers at Harvard University are studying how forests respond to drought, giving us insight on how ecosystems can endure a changing environment.</p><p>It may seem like efforts to combat this catastrophic issue are not connected, but in reality they form a collective fight for a safe future. Every one of these breakthroughs brings us closer to overcoming the challenge of drought.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*14GFZoQN36dixaMC" /><figcaption>Image by Tumisu from Pixabay</figcaption></figure><p><a href="https://phys.org/news/2024-09-scientists-ai-future-drought-conditions.html">Scientists develop AI models able to predict future drought conditions with high accuracy</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d4bc8b40194d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Forecasting Droughts with Decision Trees and Random Forests]]></title>
            <link>https://medium.com/@tanush.siva/forecasting-droughts-with-decision-trees-and-random-forests-bb2589a760fb?source=rss-e341653a8265------2</link>
            <guid isPermaLink="false">https://medium.com/p/bb2589a760fb</guid>
            <category><![CDATA[computer-science]]></category>
            <category><![CDATA[high-school]]></category>
            <category><![CDATA[random-forest]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[quantum-computing]]></category>
            <dc:creator><![CDATA[Tanush Siva]]></dc:creator>
            <pubDate>Sun, 22 Dec 2024 02:01:47 GMT</pubDate>
            <atom:updated>2024-12-22T02:01:47.181Z</atom:updated>
            <content:encoded><![CDATA[<h4>Stopping Drought and Famine with Quantum Computing</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/612/1*pj_TE6NB-j1GJUwiArjE4Q.jpeg" /></figure><p>Droughts are among the most devastating natural disasters, wreaking havoc on communities, economies, and the environment. What if there was a quantum way we could predict them before they hit? By leveraging the power of <strong>decision tree classifiers</strong> and <strong>random forests</strong>, we can stay one step ahead of one of nature’s most dangerous challenges.</p><h3>What are Decision Tree Classifiers?</h3><p>Decision trees, commonly used hierarchical tree-like structures in machine learning, make decisions, predictions, and classifications given a dataset.</p><h4>The Structure of Decision Trees</h4><p>Decision trees are made up of root, internal, and leaf nodes, as well as branches.</p><ol><li><strong>Root Node</strong>: The initial node of the decision tree, containing the entire, original dataset.</li><li><strong>Internal Node</strong>: Represent test conditions used to split the tree. When classifying a sample, the data is run through the test attribute. Depending on the outcome, the sample will move on to a different section of the tree.</li><li><strong>Leaf Node</strong>: Represent the final prediction. At this point, no further splitting will occur.</li><li><strong>Branch</strong>: Represent the outcome of a sample at an internal node.</li></ol><h4>Constructing Decision Trees</h4><p>So, how are decision trees constructed?</p><ol><li><strong>First, initialize a set of parameters/constraints for the tree to follow. </strong>Some commonly used constraints are: <strong><br></strong>- <strong>Max depth </strong>of the tree<br>- The <strong>minimum number </strong>of samples at a node for a split to occur<br>- The <strong>number of features </strong>the decision tree should consider.</li><li><strong>Next, construct the decision tree by recursively splitting the data set.</strong><br>- <strong>Check the stop criteria</strong>: Have we violated any of the provided constraints at the current node?<br>- <strong>Find the best split</strong>: Given the number of features to consider as well as a metric like Gini impurity, entropy, or information gain, select a way to split the data.<br>- <strong>Divide </strong>the samples at the node into two using the best split.<br>- <strong>Repeat </strong>this process until the stop criteria is met. At this point, a leaf node should be created. The most common label at the node will serve as the classification output.</li><li><strong>What are the common metrics for splitting?<br>- Entropy</strong>: A measurement of the amount of impurity/uncertainty at the node.<br>- <strong>Gini Impurity</strong>: A measurement which describes how often a randomly chosen sample at the node would be labeled incorrectly (depending on the label distribution at the node).<br>- <strong>Information Gain</strong>: A measurement of how much the entropy or Gini impurity at a node reduces because of a split.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/0*MVKOh6R9nAPe7mmY.png" /></figure><h3>What are Random Forest Models?</h3><p>Random forest models take a random sample of the given data and create an ongoing set of decision trees. On their own, decision trees are very sensitive towards the data set. Oftentimes, they are prone to bias and overfit data, especially as the complexity of the data set increases. This reduces the accuracy of prediction for certain samples.</p><p>By creating a random forest model, the data overfitting commonly induced by classical decision trees are reduced. The random feature selection ensures the same data is not used for every tree, making the model less sensitive and reducing the correlation between the trees.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*-MqsXgXcDpXvwuFN.png" /></figure><h3>Issues With Classical Models</h3><p>Nonetheless, both decision trees and random forest rely on classical computing. By leveraging the power of quantum computing, quantum random forests can be created, offering exponential speed-ups over classical machine learning models.</p><p>Stay tuned for part two of this article, which delves more into classical machine learning models and introduces the concept of quantum random forests!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bb2589a760fb" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Acing the IBM Quantum Certification Exam! (Part 2)]]></title>
            <link>https://medium.com/@tanush.siva/acing-the-ibm-quantum-certification-exam-part-2-86dba3084890?source=rss-e341653a8265------2</link>
            <guid isPermaLink="false">https://medium.com/p/86dba3084890</guid>
            <category><![CDATA[high-school]]></category>
            <category><![CDATA[qiskit]]></category>
            <category><![CDATA[quantum-computing]]></category>
            <category><![CDATA[computer-science]]></category>
            <category><![CDATA[ibm-certification]]></category>
            <dc:creator><![CDATA[Tanush Siva]]></dc:creator>
            <pubDate>Sun, 13 Oct 2024 00:42:01 GMT</pubDate>
            <atom:updated>2024-10-13T00:42:01.712Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*e58Jk2ngtyWN6Wcxhrjqfw.jpeg" /></figure><p>By now, you should have a solid grasp of the basics of quantum computing. The principle of quantum superposition and visualizing qubits through Bloch spheres should seem more familiar to you. Still, as you get ready to ace IBM’s Qiskit Associate Developer Certification Exam, you might be asking yourself:</p><ol><li><strong>What does it really take to pass the test?</strong></li><li><strong>What skills are the most important to learn?</strong></li></ol><p>In this second part of your guide to crack the IBM Quantum Computing Exam, I’ll be uncovering more <strong>insider tips </strong>about the test and <strong>essential skills</strong> necessary to pass the exam!</p><h4><strong>What is the IBM Qiskit Associate Developer Certification exam?</strong></h4><p>An IBM Qiskit Developer is an individual who demonstrates fundamental knowledge of quantum computing concepts and is able to express them using the Qiskit software. By passing the test, you prove your capability to use the Qiskit framework and Python programming to create and execute quantum circuits on IBM Quantum simulators.</p><h4><strong>Pauli and Hadamard Gates:</strong></h4><p><strong>Gates</strong> are used to perform logical operations on a set of inputs to produce a single output. These gates form the basis of <strong>quantum circuits</strong>, allowing us to change the states of qubits.</p><p><strong>Pauli Gates:</strong></p><p>In quantum computing, the<strong> Pauli-X gate</strong> is equivalent to the classical <strong>NOT-gate</strong>, flipping the state of the qubit from |0⟩ or |1⟩. If the state were to lie on the Z-axis, the phase of the qubit would become 2π — the original phase. This gate will not change the phase of a qubit in the phase of 0 or π.</p><p>The <strong>Pauli-Y gate </strong>is somewhat similar to the X-Gate. This gate changes the state of |0⟩ to <em>i</em>|1⟩, and |1⟩ to -<em>i</em>|0⟩. On the Z-axis, the Y-gate will also change the phase of the qubit to be 2π — the original phase. But this gate will not change the phase of a qubit in the phase of π/2 or 3π/2.</p><p>The <strong>Pauli-Z gate </strong>is known as the phase-flip gate, since it rotates the phase of the qubit by π radians when on the Z-axis. This gate will keep the probabilities of measuring |0⟩ or |1⟩ the same but can flip the states of |+⟩ to |-⟩ and vice versa.</p><p><strong>Hadamard Gate:</strong></p><p>The Hadamard gate is able to put a qubit into superposition, making it an equal probability to measure |0⟩ or |1⟩ (50%). Specifically, when applying a Hadamard gate on |0⟩, the state is turned into 1/sqrt(2)(|0⟩ + |1⟩) (|+⟩) and |1⟩ is turned into 1/sqrt(2)(|0⟩ — |1⟩) (|-⟩). Because of this, two Hadamard gates will always undo themselves.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/850/0*0mdjExhEYBsh1v1K" /></figure><h4><strong>Qiskit Visualization Methods</strong></h4><p>Now, how do we actually visualize the states of qubits and the outputs of quantum circuits? Enter the world of Qiskit visualization methods!</p><p><strong>Counts Visualizations:</strong></p><ol><li>plot_histogram(): Plot a histogram of input counts data</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/947/1*7ZFp1i4ihqCH8lwAB7jnEQ.png" /></figure><p><strong>State Visualizations:</strong></p><p>plot_bloch_vector(): Plot the Bloch sphere.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/821/1*nel5y5zXfuZOjMy-FQajGQ.png" /></figure><p>plot_bloch_multivector(): Plot a Bloch sphere for each qubit</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/861/1*r7I0birRUDCOle_27kLGjQ.png" /></figure><p>plot_state_city(): Plot the cityscape of a quantum state.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Z23j0sXQawX3NAjq" /></figure><p>plot_state_hinton(): Plot a hinton diagram for the density matrix of a quantum state.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/0*Eo_zspRrY9Nt5LAw" /></figure><p><strong>Device Visualizations:</strong></p><p>plot_gate_map(): Plots gate map of a device.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/143/0*uCTNXPUhYYZA_Mti" /></figure><p>plot_error_map(): Plots error map of a backend.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*0hvdmySesZ-6Ywek" /></figure><p>plot_coupling_map(): Plots the coupling map of qubits.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/217/0*_R6myvtAqnFGEehg" /></figure><p><strong>Circuit Visualization:</strong></p><p>draw(): Draws quantum circuit.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/858/1*sDm-Upeg4hUtICCf2u9Iig.png" /></figure><p>In this second part of the guide, we delved into the secrets of <strong>Pauli and Hadamard Gates</strong> and mastered <strong>Qiskit visualization techniques</strong>. With these skills, you are now even more prepared to conquer the <strong>IBM Qiskit Associate Developer Certification Exam</strong>!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*oKY_rf8uL4gD9k-mBg-E2A.jpeg" /></figure><p><a href="https://www.ibm.com/training/certification/ibm-certified-associate-developer-quantum-computation-using-qiskit-v02x-C0010300">IBM Certified Associate Developer - Quantum Computation using Qiskit v0.2X - IBM Training - Global</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=86dba3084890" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A Guide to the IBM Quantum Certification Exam: Unlocking Your Quantum Potential!]]></title>
            <link>https://medium.com/@tanush.siva/a-guide-to-the-ibm-quantum-certification-exam-unlocking-your-quantum-potential-a1e3625dfa50?source=rss-e341653a8265------2</link>
            <guid isPermaLink="false">https://medium.com/p/a1e3625dfa50</guid>
            <category><![CDATA[qiskit]]></category>
            <category><![CDATA[high-school]]></category>
            <category><![CDATA[quantum]]></category>
            <category><![CDATA[programming]]></category>
            <category><![CDATA[quantum-computing]]></category>
            <dc:creator><![CDATA[Tanush Siva]]></dc:creator>
            <pubDate>Sun, 22 Sep 2024 16:04:21 GMT</pubDate>
            <atom:updated>2024-09-22T16:04:21.433Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/454/1*7OuVGvKVWE6TD8kscgshKg.gif" /></figure><p>Quantum computing is set to revolutionize technology as a whole, unlocking the power to solve problems that today’s supercomputers can barely comprehend. While this may seem a few years away, the quantum revolution is happening now, at this very moment!</p><p>Your first step into this <strong>quantum future </strong>lies within the IBM Qiskit Associate Developer Certification exam.</p><p>I can help you <strong>pass </strong>this exam, so you too can join the <strong>quantum world</strong>!</p><h3><strong>What is the IBM Qiskit Developer Certification Exam?</strong></h3><p>An IBM Qiskit Developer is an individual who demonstrates fundamental knowledge of quantum computing concepts and is able to express them using the Qiskit software. By passing the test, you prove your capability to use the Qiskit framework and Python programming to create and execute quantum circuits on IBM Quantum simulators.</p><h4><strong>What are the skills needed to pass the exam?</strong></h4><ul><li>Knowledge of creating, executing, and visualizing the results quantum of circuits using the online IBM Quantum Composer.</li><li>Knowledge of developing examples that highlight features of the Qiskit SDK, such as those found in the online IBM Quantum Lab.</li><li>Knowledge of modeling quantum states and evolution with complex vectors and matrices.</li><li>Knowledge of Pauli matrices.</li><li>Knowledge of quantum state measurement probabilities.</li><li>Familiarity with common circuits (such as Bell States)</li></ul><h4><strong>How is the exam structured?</strong></h4><p>The developer exam has 60 questions, with a time limit of 90 minutes. To pass the exam, you need to get 44 out of 60 questions correct (a score of 73%).</p><p>The <strong>topics </strong>on the exam, along with their <strong>weightage</strong>, are as follows:</p><ul><li>Section 1: Perform Operations on Quantum Circuits (47%)</li><li>Section 2: Executing Experiments (3%)</li><li>Section 3: Implement BasicAer: Python-based Simulators (3%)</li><li>Section 4: Implement Qasm (1%)</li><li>Section 5: Compare and Contrast Quantum Information (10%)</li><li>Section 6: Return the Experiment Results (7%)</li><li>Section 7: Use Qiskit Tools (1%)</li><li>Section 8: Display and Use System Information (3%)</li><li>Section 9: Construct Visualizations (19%)</li><li>Section 10: Access Aer Provider (6%)</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DduSdRFkWW3PHcoaQTSeZQ.jpeg" /></figure><h3><strong>Important Concepts in the Exam</strong></h3><h4><strong>Quantum Superposition</strong></h4><p>The key difference between classical and quantum computing lies within the principle of <strong>quantum superposition</strong>. Normal, classical computers use classical bits, which can be in either a state of 0 or 1. Quantum computers use <strong>qubits</strong>, which can represent a state of 0 or 1, but also a “mix” between these two states.</p><p>To understand this concept, picture a coin flip. In classical computing, after flipping the coin, it can either be heads (a state of 0) or tails (a state of 1). In quantum computing, while the coin is spinning in the air, it is in a <strong>superposition </strong>of both heads and tails. During this time, you are <strong>unable to determine</strong> the state the coin is in. Only after it lands (that is, you <strong>measure the qubit</strong>), can you determine its state.</p><h4><strong>The Bloch Sphere</strong></h4><p>The most common way to visualize a qubit in superposition is through the <strong>Bloch sphere</strong>. Without performing any quantum operations, a qubit can hold one of two states: |0⟩ or |1⟩ (the north and south pole on the Bloch sphere. When in superposition, a qubit can hold a combination of these states, represented as other points on the surface of the Bloch sphere.</p><p>This geometric representation of qubits helps to visualize the probabilities of measuring the qubit in either state.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/374/0*5i2nUE_1eRPK7hHn" /><figcaption>Representation of a qubit after a Hadamard gate is applied.</figcaption></figure><p>As the groundbreaking field of quantum computing continues to grow, it promises to solve complex problems we once thought to be impossible. Get ready to explore the countless opportunities that quantum computing can offer. Remember: the quantum future starts now!</p><p>Stay tuned for Part Two of this article, where I will be giving more <strong>insider tips</strong> about passing the exam and taking your first steps to begin your <strong>quantum computing journey</strong>!</p><p><a href="https://www.ibm.com/training/certification/ibm-certified-associate-developer-quantum-computation-using-qiskit-v02x-C0010300">IBM Certified Associate Developer - Quantum Computation using Qiskit v0.2X - IBM Training - Global</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*f899PNemFNck2mGJBxdz0Q.jpeg" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a1e3625dfa50" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Quantum Computing: The Key to Conquering Drought and Global Famine]]></title>
            <link>https://medium.com/@tanush.siva/quantum-computing-the-key-to-conquering-drought-and-global-famine-6922fcc88997?source=rss-e341653a8265------2</link>
            <guid isPermaLink="false">https://medium.com/p/6922fcc88997</guid>
            <category><![CDATA[computer-science]]></category>
            <category><![CDATA[high-school]]></category>
            <category><![CDATA[optimization]]></category>
            <category><![CDATA[quantum-computing]]></category>
            <category><![CDATA[machine-learning]]></category>
            <dc:creator><![CDATA[Tanush Siva]]></dc:creator>
            <pubDate>Wed, 28 Aug 2024 00:06:13 GMT</pubDate>
            <atom:updated>2024-08-28T00:06:13.732Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/702/1*cnFJiEQOO159Czs0WsUunA.gif" /></figure><h4>Stopping Drought and Famine with Quantum Computing</h4><p>Over much of the world drought is spreading, devastating water resources, agriculture and compromising human health. Catastrophe looms unless we find a way to detect and prevent the drying up of our world.</p><p>Hi, my name is Tanush Siva, and I am 16. I have recently become an IBM Certified Associate Developer for quantum computation using Qiskit. With the power of quantum computing, I plan to end drought and famine worldwide.</p><p>It could and can happen!</p><p>My project predicts and protects areas that are, or will be affected by drought, via a <strong>Quantum Machine Learning </strong>(QML) model, in the form of a <strong>random forest </strong>(QRF), and a <strong>Quadratic Unconstrained Binary Optimization </strong>(QUBO) model.</p><h4>The Quantum Random Forest</h4><p>To predict occurrences of drought, the QML model will be trained on data about the conditions before a drought occurs, when it starts to occur, and during the drought itself. Specifically, this will take the form of a <strong>Quantum Random Forest </strong>(QRF), a machine learning algorithm used to make classifications of data. During training time, it constructs multiple decision trees made up of <strong>bootstrapped datasets, </strong>created from random samples of the original data. The model averages the predictions of the individual trees to reach a single, accurate result.</p><p>Each tree the QRF creates is known as a <strong>quantum decision tree </strong>(QDT), which has a similar structure to a binary decision tree. Each vertex (or <strong>node</strong>) in the QDT is classified into one of two types:</p><ol><li><strong>Split</strong> node: Responsible for separating the data point into two subsets which continue down the tree, based on a <strong>quantum split function</strong>.</li><li><strong>Leaf </strong>node: The terminal point at a branch of the QDT which determines its classification output. A leaf node is classified as a leaf node if any of the following criteria are met: <strong>A. </strong>The training instances supplied are of a single class/category. <strong>B.</strong> The number of data points is smaller than a user defined minimum value. <strong>C.</strong> The node is at the maximum depth of the tree.</li></ol><p>This method has an <strong>underlying issue</strong>: the performance of the split function. In this case, the parameters of the split function need to be manually selected or algorithmically determined during the model’s training phase. Additionally, for the split function to be effective, it must generalize well; otherwise, repeated splitting can lead to <strong>data overfitting</strong>.</p><p>It is advantageous to use <strong>Nyström Kernel Estimation </strong>as the split function, which maps the data to a higher dimensional quantum feature space to make distinguishing between classes easier.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kl3jaDvPL1AEgp7kv1X8-Q.jpeg" /></figure><h4>The QUBO Model</h4><p>Using the output of the QRF, regions to be affected by drought will be made clear. This information will be used in a <strong>QUBO </strong>model, which will equitably allocate water among regions through the use of an <strong>objective function</strong> with <strong>constraints</strong>.</p><p>A QUBO model expresses an optimization problem into <strong>binary quadratic form</strong> and searches for an optimal solution by evaluating different combinations of <strong>binary variables</strong>. It aims to find a set of binary variables that either maximizes or minimizes the overall QUBO value, which is the sum of the objective and constraint functions.</p><p>When determining water allocation, the objective function prioritizes regions with the highest likelihood of drought — which is determined from the QRF. The program loops through all distributions of water, with the binary variable representing if water is given to the region (1), or not (0). After evaluating each scenario, the model chooses the solution where the most regions with the highest priority values are given water.</p><p>This QUBO model also includes two key constraints. The first constraint in the model penalizes solution sets where less than one quarter of the regions are given water, while the second constraint ensures the total amount of water allocated per region does not exceed the total supply of water.</p><p>This model aims to maximize the overall QUBO value, increasing it when more regions with high priority values are given water, and decreasing it when certain constraints are not met.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*L1KSCXE9nHoA3LirE4VW-w.jpeg" /></figure><p>By leveraging the power of quantum computing, this program will ensure the accurate prediction and protection of nations against drought, preventing water wars and other resources conflicts globally!</p><h4><strong>Relevant Links:</strong></h4><ul><li><a href="https://www.ibm.com/training/certification/ibm-certified-associate-developer-quantum-computation-using-qiskit-v02x-C0010300">IBM Certified Associate Developer - Quantum Computation using Qiskit v0.2X - IBM Training - Global</a></li><li><a href="https://arxiv.org/abs/2210.02355">A kernel-based quantum random forest for improved classification</a></li><li><a href="https://arxiv.org/abs/1811.11538">A Tutorial on Formulating and Using QUBO Models</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6922fcc88997" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>