<?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 Esther Jokodola on Medium]]></title>
        <description><![CDATA[Stories by Esther Jokodola on Medium]]></description>
        <link>https://medium.com/@estherjokodola21?source=rss-c5f83b23d4c3------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*e6SxqLVvKVN37tyvVI5Kbw.jpeg</url>
            <title>Stories by Esther Jokodola on Medium</title>
            <link>https://medium.com/@estherjokodola21?source=rss-c5f83b23d4c3------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 15:12:31 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@estherjokodola21/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[I Built My First Neural Network — It Drew a Line Between Red and Blue Dots, and I Loved It]]></title>
            <link>https://medium.com/@estherjokodola21/i-built-my-first-neural-network-it-drew-a-line-between-red-and-blue-dots-and-i-loved-it-9b02d7b1dda5?source=rss-c5f83b23d4c3------2</link>
            <guid isPermaLink="false">https://medium.com/p/9b02d7b1dda5</guid>
            <category><![CDATA[short-read]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[neural-networks]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[education]]></category>
            <dc:creator><![CDATA[Esther Jokodola]]></dc:creator>
            <pubDate>Sun, 06 Apr 2025 01:10:17 GMT</pubDate>
            <atom:updated>2025-04-06T01:10:17.970Z</atom:updated>
            <content:encoded><![CDATA[<h3>I Built My First Neural Network — It Drew a Line Between Red and Blue Dots, and I Loved It</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/559/1*sWlaXJGWSt5mbof-3Q-D9A.png" /><figcaption>A visualization of the make_circle dataset</figcaption></figure><p>The task? To teach a model how to tell apart two sets of dots: red and blue. But here’s the twist — these dots weren’t randomly scattered. They were arranged in circles. Think of two rings, one inside the other. Our goal was to train a model that could look at a new dot and say, “Hey, you belong to the red circle,” or “You’re definitely a blue one.”</p><p>And guess what? We did it. The model learned to draw a line — well, more like a <em>curve</em> — that beautifully separated the two.</p><p><strong>The Dataset: Circles of Color</strong></p><p>For this project, I used a synthetic dataset — one that’s often used in binary classification tasks. The data points were in two classes: red and blue, each forming a ring (like a donut with a hole). It’s called the “make_circles” dataset in scikit-learn.</p><p>Each point had two features: its x and y coordinates. That’s it. But visually, they told a bigger story.</p><p>We split the data into <strong>training</strong> and <strong>testing</strong> sets. The training set helped the model learn, and the testing set checked how well it actually understood the pattern.</p><pre> from sklearn.datasets import make_circles<br><br> # Make 1000 samples<br> n_samples =1000<br><br> # Create circles<br> X,y = make_circles(n_samples,<br>                    noise=0.03,<br>                    random_state=42)</pre><pre># Split data into training and testing data sets<br>from sklearn.model_selection import train_test_split<br><br>X_train, X_test, y_train, y_test = train_test_split(X,<br>                                                     y,<br>                                                     test_size=0.2,<br>                                                     random_state=42)</pre><p><strong>The Model: My First Neural Network in PyTorch</strong></p><p>I built a very small neural network using <strong>PyTorch</strong> — my first ever!</p><p>Here’s what it had:</p><ul><li>An input layer (2 neurons — one for x, one for y)</li><li>A hidden layer with non-linearity using ReLU</li><li>An output layer with 1 neuron using the sigmoid activation function</li></ul><p>Why sigmoid? Because we wanted the output to be between 0 and 1 — perfect for binary classification, where we can treat anything ≥ 0.5 as one class, and &lt; 0.5 as the other.</p><pre><br># Build a model with non-linear activation functions<br>from torch import nn<br>class CircleModelV2(nn.Module):<br>  def __init__(self):<br>    super().__init__()<br>    self.layer_1 = nn.Linear(in_features=2, out_features=10)<br>    self.layer_2 = nn.Linear(in_features=10, out_features=10)<br>    self.layer_3 = nn.Linear(in_features=10, out_features=1)<br>    self.relu =nn.ReLU() # RELU IS A NON-LINEAR ACTIVATION FUNCTION<br><br>  def forward(self, x):<br>    # where should we put our non-linear functions?<br>    return self.layer_3(self.relu(self.layer_2(self.relu(self.layer_1(x)))))<br><br>model_3 = CircleModelV2().to(device)<br>model_3</pre><p><strong>Training the Model</strong></p><p>I trained the model using:</p><ul><li>Binary Cross-Entropy Loss (BCELoss)</li><li>Stochastic Gradient Descent (SGD) optimizer</li></ul><p>During training, I watched the loss go down — slowly at first, then faster. I even visualized how the decision boundary changed over time. At first, the line was messy, but eventually, it formed a neat curve that split the red and blue circles like magic.</p><pre># Setup loss and optimizer<br>loss_fn = nn.BCEWithLogitsLoss()<br>optimizer= torch.optim.SGD(model_3.parameters(),<br>                           lr=0.1)</pre><pre># Random seeds<br>torch.manual_seed(42)<br>torch.cuda.manual_seed(42)<br><br># Put all data on target device<br>X_train, y_train = X_train.to(device), y_train.to(device)<br>X_test, y_test = X_test.to(device), y_test.to(device)<br><br># Loop through data<br>epochs = 5000<br><br>for epoch in range(epochs):<br>  ### Training<br>  model_3.train()<br><br>  # 1. Forward pass<br>  y_logits = model_3(X_train).squeeze()<br>  y_pred = torch.round(torch.sigmoid(y_logits)) # logits -&gt; prediction probabilities -&gt; prediction labels<br><br>  # 2. Calculate the loss<br>  loss = loss_fn(y_logits, y_train) # BCEWithLogitsLoss (takes in logits as first input)<br>  acc = accuracy_fn(y_true=y_train,<br>                    y_pred=y_pred)<br><br>  # 3. Optimizer zero grad<br>  optimizer.zero_grad()<br><br>  # 4. Loss backward<br>  loss.backward()<br><br>  # 5. Step the optimizer<br>  optimizer.step()<br><br>  ### Testing<br>  model_3.eval()<br>  with torch.inference_mode():<br>    test_logits = model_3(X_test).squeeze()<br>    test_pred = torch.round(torch.sigmoid(test_logits))<br><br>    test_loss = loss_fn(test_logits, y_test)<br>    test_acc = accuracy_fn(y_true=y_test,<br>                           y_pred=test_pred)<br><br>  # Print out what&#39;s this happenin&#39;<br>  if epoch % 100 == 0:<br>    print(f&quot;Epoch: {epoch} | Loss: {loss:.4f}, Acc: {acc:.2f}% | Test Loss: {test_loss:.4f}, Test Acc: {test_acc:.2f}%&quot;)</pre><p><strong>Visualizing the Decision Boundary</strong></p><p>One of the best parts of this project was the visualization. After training, I plotted the decision boundary: the model’s best guess at where the red ends and the blue begins.</p><p>It wasn’t a straight line — because the data wasn’t linearly separable. It was a beautiful curve, shaped by the learning process itself.</p><pre># Plot decision boundaries<br>plt.figure(figsize=(12,6))<br>plt.subplot(1,2,1)<br>plt.title(&quot;Train&quot;)<br>plot_decision_boundary(model_3, X_train, y_train)<br>plt.subplot(1,2,2)<br>plt.title(&quot;Test&quot;)<br>plot_decision_boundary(model_3, X_test,y_test)</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/993/1*BAZQiDOPgtYbiXfyjYtwTg.png" /><figcaption>A visualization of the train dataset and test dataset after being trained by the model</figcaption></figure><p><strong>Why This Meant So Much to Me</strong></p><p>I know this wasn’t some giant AI breakthrough. But to me, it meant everything.</p><p>It was the moment I saw theory turn into something real. The math, the code, the colors — all came together. I got to <em>see</em> my model learning, and that made me fall in love with machine learning all over again.</p><p><strong>Final Thoughts</strong></p><p>This was a small project, yes — but it laid a solid foundation. It taught me how to:</p><ul><li>Work with data</li><li>Build and train a neural network using PyTorch</li><li>Visualize the learning process</li><li>Appreciate even the tiniest model victories</li></ul><p>I’m just getting started, but if you’re also new to machine learning, let me tell you: your first model doesn’t need to be huge. Just make it <em>yours</em>.</p><p>Thanks for reading! If you’d like to check out the code or visuals, feel free to visit my <a href="https://github.com/Estherjokodola/End-to-End-Circular-Boundary-Classification-using-Neural-Network/blob/main/Neural_Boundary_Classification.ipynb">github</a>.</p><p>Follow me here and my other socials for more content.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9b02d7b1dda5" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[“Who am I?”]]></title>
            <link>https://medium.com/@estherjokodola21/who-am-i-1c7e57ca9f93?source=rss-c5f83b23d4c3------2</link>
            <guid isPermaLink="false">https://medium.com/p/1c7e57ca9f93</guid>
            <category><![CDATA[self-awareness]]></category>
            <category><![CDATA[life]]></category>
            <category><![CDATA[self-care]]></category>
            <category><![CDATA[self-reflection]]></category>
            <category><![CDATA[self-improvement]]></category>
            <dc:creator><![CDATA[Esther Jokodola]]></dc:creator>
            <pubDate>Fri, 29 Mar 2024 13:34:32 GMT</pubDate>
            <atom:updated>2024-03-29T13:34:32.835Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="Discovery" src="https://cdn-images-1.medium.com/max/735/1*z9580JHhSdHJbFEuvUHzbQ.jpeg" /></figure><p>I am a 20-year-old, still figuring things out. I want to start writing to document something — anything — that’ll help the next person going through the mental difficulties that I face on a day-to-day life.</p><p>Nobody talks about how you get over depression. Nobody asks an obese individual why they want a regular person’s body or how they became obese in the first place. They are always judged based on a place of belonging, as if they fit in the lowest of hierarchy.</p><p>I am not writing this because I’m obese. I’m not. But I am an introverted emotional being who lacks self-will. They refer to us as people who lack confidence. There’s one thing people fail to understand: we are literally all in the wheel of life, seeing this as an issue and working our hardest to get out of this. Being an introverted emotional being is not a problem; it’s a gift. God (The universe) made me different from the regular person in various ways.</p><p>I am mostly referred to as immature when I decide to be fully myself. And when I ask why, they say I’m being myself to the wrong people. Then I go on to be myself to the wrong people and keep getting criticized for it, until I couldn’t tell who I was anymore. I can’t tell who I am anymore, and that’s on me. I let people who were undeserving of my raw energy, my peace, my aura, get into my life and rip me apart.</p><p>Now I am on a real quest to discover myself. I hope you resonate with some part of this and choose to find yourself also. I have decided to document both my happy and sad moments and also my reflection on them. I am not giving up on myself. I hope you won’t also. Let’s go on this self-discovery journey together. Peace and Love.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1c7e57ca9f93" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>