initial projects page

Author: csmartinez22

my-app/src/pages/Projects.css

@@ -0,0 +1,14 @@
+.header {
+
+}
+
+.text {
+    color: rgb(154, 157, 158);
+}
+
+.projects {
+    display: flex;
+    flex-direction: column;
+    justify-content: center;
+    align-items: center;
+}

my-app/src/pages/Projects.js

@@ -1,5 +1,136 @@
+import './Projects.css'
+
 const Projects = () => {
-  return <h1>Personal Projects</h1>;
+  return (
+    <div className="projects">
+      <h1 className="title">
+        Projects
+      </h1>
+      <h1 className="title">
+        Puerto Rico
+      </h1>
+      <div className="body">
+        Below is an informal write where I used machine learning to identify which municipalities in Puerto Rico are at risk of gentrification.
+      </div>
+      <h2 className="header">
+        Preface
+      </h2>
+      <div className="body">
+      I chose Puerto Rico as my focus for my project because mainland Americans often forget it. 
+      In a 2017 study, a poll conducted by Morning Consult found that 46% of Americans did not know 
+      that Puerto Ricans are considered U.S. citizens. I knew I wanted to work on gentrification as
+      someone who has lived in public housing for most of my life and saw the city of Somerville
+      destroy the public housing complex for mixed-income housing. I searched for existing projects
+      that focused on identifying gentrification in the U.S., and every project I found
+      concentrated on mainland U.S. major cities, excluding Puerto Rico. By combining both topics,
+      I aim to raise awareness of displaced Puerto Ricans, create a tool that lawmakers can use
+      to see which communities are most vulnerable to being displaced, and encourage the
+      artificial intelligence community to look at tackling more diverse issues and datasets.
+      In addition to identifying the problem, I think it’s crucial to understand how
+      Puerto Rico became at risk of gentrification to address it. Under former Puerto Rican
+      governor Luis Fortuño, two controversial acts, Act 22 and Act 60, were passed that
+      encouraged wealthy foreigners to move to Puerto Rico through tax incentives.
+      The government has been unsuccessful in ensuring new residents meet the 
+      requirements in exchange for tax incentives, allowing foreigners to take 
+      advantage of the two acts.
+      </div>
+      <h2 className="header">
+        Methodology
+      </h2>
+      <div className="body">
+      I went through the available tables on the United States Census Bureau’s website to look for 
+      data tables under Puerto Rico that I thought would be relevant to finding which municipalities
+      are at risk of or already gentrified. In total, I compiled nine tables from the census website
+      with data from 2021, including demographic and housing estimates, age and sex, geographic
+      mobility by selected characteristics, households and families, educational attainment, mean
+      income, demographic characteristics of occupied housing units, financial characteristics,
+      physical housing characteristics, and race. After compiling and cleaning the data, I 
+      considered the best approach to the problem. I initially viewed unsupervised learning as
+      the best way to approach the problem since I didn’t have access to any labeled data. I 
+      chose the K Means Clustering model and attempted to train it but found that the clusters
+      the model created seemed arbitrary and didn’t indicate any pattern. After researching 
+      K Means Clustering, I discovered that the algorithm struggles with high-dimensional data,
+      making it unideal for my dataset.
+      </div>
+      <div className="body">
+      To make K Means Clustering work, I reduced the number of tables used to two and found the model 
+      was picking up on a pattern. I looked into the categories the model created and noticed it was 
+      grouping the municipalities with major cities together. It grouped San Juan, Ponce, Caguas, and 
+      Bayamon. To verify these results, I looked at articles to find some municipalities that are 
+      currently dealing with gentrification and found that my model misclassified many municipalities 
+      dealing with gentrification as not at risk. After experimenting with the parameters and getting 
+      the same results back, I decided to experiment with the algorithm. I went with the Spectral 
+      Clustering algorithm, which often outperforms K Means Clustering. After training the model on 
+      both the complete and subsets of the dataset, I found similar results to K Means Clusters, 
+      where it would group all the municipalities with major cities.
+      </div>
+      <div className="body">
+      After being unsuccessful with two unsupervised learning algorithms, I assumed my only other option 
+      would be to use a supervised learning algorithm. I researched which algorithm would work best if 
+      given a small amount of labeled data. I came across the idea of semi-supervised learning, a 
+      middle ground between unsupervised and supervised. Semi-supervised learning, also known as weak 
+      supervision, can learn using large amounts of unlabeled data paired with labeled data. 
+      Semi-supervised learning circumvents some common issues associated with getting a dataset for 
+      supervised learning, such as being expensive and time-consuming. The approach requires less 
+      human oversight while still benefiting from a higher accuracy. 
+      </div>
+      <div className="body">
+      To create a small amount of labeled data, I searched for articles and personal posts about 
+      gentrifying parts of Puerto Rico. I found posts on Reddit of users describing the gentrification 
+      they’re witnessing while also categorizing the municipalities people moved to after being 
+      displaced as less likely to be gentrified. After labeling a small subset of data, I looked at 
+      the available unsupervised learning algorithms. I decided to use the SelfTrainingClassifier 
+      from Scikit Learn, which is based on Yarowsky’s algorithm and allows for a supervised classifier 
+      to work like a semi-supervised classifier. According to Scikit Learn, “It does this by 
+      iteratively predicting pseudo-labels for the unlabeled data and adding them to the training set.”  For my base supervised learning algorithm to pair with the SelfTrainingClassifer, I used the Support Vector Classification (SVC) algorithm.
+      </div>
+      <h2 className="header">
+        Results
+      </h2>
+      <div className="body">
+      After training the algorithm and creating a model to predict all municipalities in Puerto Rico, it 
+      found that out of the 77 municipalities in Puerto Rico, all but 11 are classified as gentrifying 
+      or at risk of gentrification. The following municipalities are considered low risk for 
+      gentrification: 
+      </div>
+      <ul>
+        <li>Bayamón </li>
+        <li>Caguas </li>
+        <li>Canóvanas </li>
+        <li>Carolina </li>
+        <li>Guaynabo </li>
+        <li>Gurabo </li>
+        <li>Juncos </li>
+        <li>Santa Isabel</li>
+        <li>Toa Alta</li>
+        <li>Toa Baja</li>
+        <li>Trujillo Alto</li>
+      </ul>
+      <div className="body">
+      Below is a map of the results of the model’s prediction for every municipality.
+      </div>
+      <h2 className="header">
+        Future Research
+      </h2>
+      <div className="body">
+      To improve my project, I’d like to ensure the initial labels I used for the SelfTrainingClassifier 
+      are as accurate as possible. For my initial labels, I researched personal stories of individuals 
+      facing gentrification in their neighborhoods. Although these stories are helpful, there are 
+      plenty of other Puerto Ricans whose stories may not have appeared in my research or had the 
+      opportunity to share their story. To ensure my labeled data is more accurate, I’d like to work 
+      with a sociologist or nonprofit that focuses on gentrification in Puerto Rico to verify my 
+      initial labels and results. I’d also be interested in conducting interviews with residents of 
+      affected municipalities or conducting a survey to see how accurate my model is.
+      </div>
+      <div className="body">
+      Another way I could improve my project would be by combining the data I collected from the U.S. Census 
+      with housing data from Zillow or AirBnB. One of the limitations of using census data is that it 
+      isn’t up to date with current events. The data I collected is from 2021, so the status of 
+      gentrification in Puerto Rico could look different now compared to back then. Using current 
+      housing data could circumvent that issue.
+      </div>
+    </div>
+  );
 };
 
 export default Projects;