123

index.html

@@ -307,18 +307,20 @@ <h1 class="f3 fw1 athelas mt0 lh-title">
 
         <div class="pr3-ns mb4 mb0-ns w-100 w-40-ns">
           <a href="https://andysucao.github.io/Andy_Portfolio/post/project-4/" class="db grow">
-            <img src="https://andysucao.github.io/Andy_Portfolio/images/project-1-1b.jpg" class="img" alt="image from Project 4: Image Classification">
+            <img src="https://andysucao.github.io/Andy_Portfolio/images/projects-4-1.jpg" class="img" alt="image from Project 4: Image Classification and Explainable Artificial Intelligence">
           </a>
         </div>
 
       <div class="blah w-100 w-60-ns pl3-ns">
         <h1 class="f3 fw1 athelas mt0 lh-title">
           <a href="https://andysucao.github.io/Andy_Portfolio/post/project-4/" class="color-inherit dim link">
-            Project 4: Image Classification
+            Project 4: Image Classification and Explainable Artificial Intelligence
             </a>
         </h1>
         <div class="f6 f5-l lh-copy nested-copy-line-height nested-links">
-
+          1. Project Overview In this project, we will build a model for image classification and understand how it works.
+In the first part, we will develop a convolutional neural network (CNN) model for food image classificaton. We will also apply t-distributed Stochastic Neighbor Embedding (t-SNE) technique on the output of different layers to visualize learned visual representations of the CNN model.
+In order to understand how the model works, we will employ four popular Explainable AI approaches in the second part, including (1) Saliency map, (2) Smooth gradient, (3) Lime package, and (4) Integrated gradients.
         </div>
           <a href="https://andysucao.github.io/Andy_Portfolio/post/project-4/" class="ba b--moon-gray bg-light-gray br2 color-inherit dib f7 hover-bg-moon-gray link mt2 ph2 pv1">read more</a>
 
@@ -420,42 +422,11 @@ <h1 class="f3 fw1 athelas mt0 lh-title">
         </h1>
         <div class="f6 f5-l lh-copy nested-copy-line-height nested-links">
           1. Project Overview In this project, we will build a linear regression model to help people better understand the factors that can affect the price of a house in Beijing.
-The main sections of this article are: (1) Exploratory Data Analysis (EDA), (2) Developing Linear Regression model for predicting real estate price, and (3) Use that model to make predictions. The Pytyhon Notebook containing the complete model development process can be found at Google Drive.
+The main sections of this article are: (1) Exploratory Data Analysis (EDA), (2) Developing Linear Regression model for predicting real estate price, and (3) Use that model to make predictions.
+The Pytyhon Notebook containing the complete model development process and the data used in this project can be found at Google Drive.
         </div>
           <a href="https://andysucao.github.io/Andy_Portfolio/post/project-1/" class="ba b--moon-gray bg-light-gray br2 color-inherit dib f7 hover-bg-moon-gray link mt2 ph2 pv1">read more</a>
 
-      </div>
-    </div>
-  </div>
-</article>
-
-          </div>
-
-          <div class="relative w-100 mb4">
-
-<article class="bb b--black-10">
-  <div class="db pv4 ph3 ph0-l no-underline dark-gray">
-    <div class="flex flex-column flex-row-ns">
-
-
-
-        <div class="pr3-ns mb4 mb0-ns w-100 w-40-ns">
-          <a href="https://andysucao.github.io/Andy_Portfolio/post/project-0/" class="db grow">
-            <img src="https://andysucao.github.io/Andy_Portfolio/images/project-1-1b.jpg" class="img" alt="image from Project 0: Template &amp; Misc">
-          </a>
-        </div>
-
-      <div class="blah w-100 w-60-ns pl3-ns">
-        <h1 class="f3 fw1 athelas mt0 lh-title">
-          <a href="https://andysucao.github.io/Andy_Portfolio/post/project-0/" class="color-inherit dim link">
-            Project 0: Template &amp; Misc
-            </a>
-        </h1>
-        <div class="f6 f5-l lh-copy nested-copy-line-height nested-links">
-
-        </div>
-          <a href="https://andysucao.github.io/Andy_Portfolio/post/project-0/" class="ba b--moon-gray bg-light-gray br2 color-inherit dib f7 hover-bg-moon-gray link mt2 ph2 pv1">read more</a>
-
       </div>
     </div>
   </div>

index.xml

@@ -44,12 +44,14 @@
     </item>
 
     <item>
-      <title>Project 4: Image Classification</title>
+      <title>Project 4: Image Classification and Explainable Artificial Intelligence</title>
       <link>https://andysucao.github.io/Andy_Portfolio/post/project-4/</link>
       <pubDate>Sun, 20 Aug 2023 06:03:00 -0400</pubDate>
 
       <guid>https://andysucao.github.io/Andy_Portfolio/post/project-4/</guid>
-      <description></description>
+      <description>1. Project Overview In this project, we will build a model for image classification and understand how it works.
+In the first part, we will develop a convolutional neural network (CNN) model for food image classificaton. We will also apply t-distributed Stochastic Neighbor Embedding (t-SNE) technique on the output of different layers to visualize learned visual representations of the CNN model.
+In order to understand how the model works, we will employ four popular Explainable AI approaches in the second part, including (1) Saliency map, (2) Smooth gradient, (3) Lime package, and (4) Integrated gradients.</description>
     </item>
 
     <item>
@@ -82,7 +84,8 @@ In the first part, we will conduct Exploratory Data Analysis (EDA). Here the Syn
 
       <guid>https://andysucao.github.io/Andy_Portfolio/post/project-1/</guid>
       <description>1. Project Overview In this project, we will build a linear regression model to help people better understand the factors that can affect the price of a house in Beijing.
-The main sections of this article are: (1) Exploratory Data Analysis (EDA), (2) Developing Linear Regression model for predicting real estate price, and (3) Use that model to make predictions. The Pytyhon Notebook containing the complete model development process can be found at Google Drive.</description>
+The main sections of this article are: (1) Exploratory Data Analysis (EDA), (2) Developing Linear Regression model for predicting real estate price, and (3) Use that model to make predictions.
+The Pytyhon Notebook containing the complete model development process and the data used in this project can be found at Google Drive.</description>
     </item>
 
     <item>
@@ -121,15 +124,6 @@ R Programming course certificate
 Getting and Cleaning Data course certificate</description>
     </item>
 
-    <item>
-      <title>Project 0: Template &amp; Misc</title>
-      <link>https://andysucao.github.io/Andy_Portfolio/post/project-0/</link>
-      <pubDate>Tue, 11 Jul 2023 08:58:08 -0400</pubDate>
-
-      <guid>https://andysucao.github.io/Andy_Portfolio/post/project-0/</guid>
-      <description> </description>
-    </item>
-
     <item>
       <title>Contact</title>
       <link>https://andysucao.github.io/Andy_Portfolio/contact/</link>

post/index.html

@@ -246,11 +246,13 @@ <h1 class="f3 near-black">
 
     <h1 class="f3 near-black">
       <a href="https://andysucao.github.io/Andy_Portfolio/post/project-4/" class="link black dim">
-        Project 4: Image Classification
+        Project 4: Image Classification and Explainable Artificial Intelligence
       </a>
     </h1>
     <div class="nested-links f5 lh-copy nested-copy-line-height">
-
+      1. Project Overview In this project, we will build a model for image classification and understand how it works.
+In the first part, we will develop a convolutional neural network (CNN) model for food image classificaton. We will also apply t-distributed Stochastic Neighbor Embedding (t-SNE) technique on the output of different layers to visualize learned visual representations of the CNN model.
+In order to understand how the model works, we will employ four popular Explainable AI approaches in the second part, including (1) Saliency map, (2) Smooth gradient, (3) Lime package, and (4) Integrated gradients.
     </div>
   </div>
 
@@ -316,27 +318,8 @@ <h1 class="f3 near-black">
     </h1>
     <div class="nested-links f5 lh-copy nested-copy-line-height">
       1. Project Overview In this project, we will build a linear regression model to help people better understand the factors that can affect the price of a house in Beijing.
-The main sections of this article are: (1) Exploratory Data Analysis (EDA), (2) Developing Linear Regression model for predicting real estate price, and (3) Use that model to make predictions. The Pytyhon Notebook containing the complete model development process can be found at Google Drive.
-    </div>
-  </div>
-
-        </div>
-
-        <div class="relative w-100 w-30-l mb4 bg-white">
-
-            <div class="mb3 pa4 mid-gray overflow-hidden">
-
-      <div class="f6">
-        July 11, 2023
-      </div>
-
-    <h1 class="f3 near-black">
-      <a href="https://andysucao.github.io/Andy_Portfolio/post/project-0/" class="link black dim">
-        Project 0: Template &amp; Misc
-      </a>
-    </h1>
-    <div class="nested-links f5 lh-copy nested-copy-line-height">
-
+The main sections of this article are: (1) Exploratory Data Analysis (EDA), (2) Developing Linear Regression model for predicting real estate price, and (3) Use that model to make predictions.
+The Pytyhon Notebook containing the complete model development process and the data used in this project can be found at Google Drive.
     </div>
   </div>
 

post/index.xml

@@ -44,12 +44,14 @@
     </item>
 
     <item>
-      <title>Project 4: Image Classification</title>
+      <title>Project 4: Image Classification and Explainable Artificial Intelligence</title>
       <link>https://andysucao.github.io/Andy_Portfolio/post/project-4/</link>
       <pubDate>Sun, 20 Aug 2023 06:03:00 -0400</pubDate>
 
       <guid>https://andysucao.github.io/Andy_Portfolio/post/project-4/</guid>
-      <description></description>
+      <description>1. Project Overview In this project, we will build a model for image classification and understand how it works.
+In the first part, we will develop a convolutional neural network (CNN) model for food image classificaton. We will also apply t-distributed Stochastic Neighbor Embedding (t-SNE) technique on the output of different layers to visualize learned visual representations of the CNN model.
+In order to understand how the model works, we will employ four popular Explainable AI approaches in the second part, including (1) Saliency map, (2) Smooth gradient, (3) Lime package, and (4) Integrated gradients.</description>
     </item>
 
     <item>
@@ -82,16 +84,8 @@ In the first part, we will conduct Exploratory Data Analysis (EDA). Here the Syn
 
       <guid>https://andysucao.github.io/Andy_Portfolio/post/project-1/</guid>
       <description>1. Project Overview In this project, we will build a linear regression model to help people better understand the factors that can affect the price of a house in Beijing.
-The main sections of this article are: (1) Exploratory Data Analysis (EDA), (2) Developing Linear Regression model for predicting real estate price, and (3) Use that model to make predictions. The Pytyhon Notebook containing the complete model development process can be found at Google Drive.</description>
-    </item>
-
-    <item>
-      <title>Project 0: Template &amp; Misc</title>
-      <link>https://andysucao.github.io/Andy_Portfolio/post/project-0/</link>
-      <pubDate>Tue, 11 Jul 2023 08:58:08 -0400</pubDate>
-
-      <guid>https://andysucao.github.io/Andy_Portfolio/post/project-0/</guid>
-      <description> </description>
+The main sections of this article are: (1) Exploratory Data Analysis (EDA), (2) Developing Linear Regression model for predicting real estate price, and (3) Use that model to make predictions.
+The Pytyhon Notebook containing the complete model development process and the data used in this project can be found at Google Drive.</description>
     </item>
 
   </channel>

post/project-1/index.html

@@ -34,7 +34,7 @@
 <meta itemprop="name" content="Project 1: Predict Real Estate Prices in Beijing">
 <meta itemprop="description" content="A linear regression based model to predict house price"><meta itemprop="datePublished" content="2023-08-20T06:00:00-04:00" />
 <meta itemprop="dateModified" content="2023-08-20T06:00:00-04:00" />
-<meta itemprop="wordCount" content="2400">
+<meta itemprop="wordCount" content="2407">
 <meta itemprop="keywords" content="Machine Learning,Linear Regression," /><meta name="twitter:card" content="summary"/>
 <meta name="twitter:title" content="Project 1: Predict Real Estate Prices in Beijing"/>
 <meta name="twitter:description" content="A linear regression based model to predict house price"/>
@@ -191,7 +191,8 @@ <h1 class="f1 athelas mt3 mb1">Project 1: Predict Real Estate Prices in Beijing<
     </header>
     <div class="nested-copy-line-height lh-copy serif f4 nested-links nested-img mid-gray pr4-l w-two-thirds-l"><h2 id="1-project-overview">1. Project Overview</h2>
 <p>In this project, we will build a linear regression model to help people better understand the factors that can affect the price of a house in Beijing.</p>
-<p>The main sections of this article are: (1) Exploratory Data Analysis (EDA), (2) Developing Linear Regression model for predicting real estate price, and (3) Use that model to make predictions. The Pytyhon Notebook containing the complete model development process can be found at <a href="https://colab.research.google.com/drive/1oismdB2dYLYyhVjcLMazenDPd9WCd5SI?usp=sharing">Google Drive</a>.</p>
+<p>The main sections of this article are: (1) Exploratory Data Analysis (EDA), (2) Developing Linear Regression model for predicting real estate price, and (3) Use that model to make predictions.</p>
+<p>The Pytyhon Notebook containing the complete model development process and the data used in this project can be found at <a href="https://drive.google.com/drive/folders/137IaDtqgBPX4F6zk6VJYEBSfibrdmhHW?usp=sharing">Google Drive</a>.</p>
 <p>           
            </p>
 <h2 id="2-exploratory-data-analysis-eda">2. Exploratory Data Analysis (EDA)</h2>

post/project-2/index.html

@@ -34,7 +34,7 @@
 <meta itemprop="name" content="Project 2: Predicitng Customer Churn for a Mobile Phone Carrier">
 <meta itemprop="description" content="Develop machine learning models to predict customer churn using Logistic Regression, Decision Tree, Ensemble methods (Random Forest, Adaboost, GBDT), and Neural Network"><meta itemprop="datePublished" content="2023-08-20T06:01:00-04:00" />
 <meta itemprop="dateModified" content="2023-08-20T06:01:00-04:00" />
-<meta itemprop="wordCount" content="1468">
+<meta itemprop="wordCount" content="1475">
 <meta itemprop="keywords" content="Machine Learning,Logistic Regression," /><meta name="twitter:card" content="summary"/>
 <meta name="twitter:title" content="Project 2: Predicitng Customer Churn for a Mobile Phone Carrier"/>
 <meta name="twitter:description" content="Develop machine learning models to predict customer churn using Logistic Regression, Decision Tree, Ensemble methods (Random Forest, Adaboost, GBDT), and Neural Network"/>
@@ -228,7 +228,7 @@ <h1 class="f1 athelas mt3 mb1">Project 2: Predicitng Customer Churn for a Mobile
 </tbody>
 </table>
 <hr>
-<p>The Pytyhon Notebook containing the complete model development process can be found at <a href="https://colab.research.google.com/drive/18STLpbYJCAAv135uwNc3Z9s8yAmy3u1A?usp=sharing">Google Drive</a>.</p>
+<p>The Pytyhon Notebook containing the complete model development process and the data used in this project can be found at <a href="https://drive.google.com/drive/folders/1huLD3Pm9pZIhcqHSl4XCoRZgQIxQ1uPj?usp=sharing">Google Drive</a>.</p>
 <p>           
            </p>
 <h2 id="2-exploratory-data-analysis-eda">2. Exploratory Data Analysis (EDA)</h2>
@@ -512,8 +512,7 @@ <h3 id="34-neural-network">3.4. Neural Network</h3>
 <p>           
            </p>
 <h2 id="4-conclusions">4. Conclusions</h2>
-<p>In this work, I have developed a series of models to predict customer churn for a mobile phone carrier. Among all the methods I tried, I would recommend
-Adaboost if model explainability is not required, because it is fast, robust, and very accurate. On the other hand, if &ldquo;black box&rdquo; model is not acceptable, then I would recommend Multivariate logistic regression with L1 regularization, because it is explainable and with careful hyperparameter tuning, very accurate prediction too.</p>
+<p>In this work, I have developed a series of models to predict customer churn for a mobile phone carrier. Among all the methods I tried, I would recommend Adaboost if model explainability is not required, because it is fast, robust, and very accurate. On the other hand, if &ldquo;black box&rdquo; model is not acceptable, then I would recommend Multivariate logistic regression with L1 regularization, because it is explainable and with careful hyperparameter tuning, very accurate prediction too.</p>
 <ul class="pa0">
 
    <li class="list">