-
Notifications
You must be signed in to change notification settings - Fork 0
/
Pig_E_Bank.html
226 lines (203 loc) · 10.8 KB
/
Pig_E_Bank.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
<!DOCTYPE HTML>
<!--
Solid State by HTML5 UP
html5up.net | @ajlkn
Free for personal and commercial use under the CCA 3.0 license (html5up.net/license)
-->
<html>
<head>
<title>Manzano Analytics</title>
<meta charset="utf-8" />
<meta name="viewport" content="width=device-width, initial-scale=1, user-scalable=no" />
<link rel="stylesheet" href="assets/css/main.css" />
<noscript><link rel="stylesheet" href="assets/css/noscript.css" /></noscript>
<link rel="icon" type="image/png" href="images/geo_flake_b&w.png">
<style>
.content2 {
background-color: rgba(0, 0, 0, 0.55); /* Dark background color */
margin-left: auto; /* Add padding to improve readability */
margin-right: auto;
width: 50%;
color: #ffffff;
border-radius: 10px; /* Add rounded corners for aesthetic appeal */
}
.hover-zoom:hover {
transform: scale(1.1); /* Increase size by 10% on hover */
}
</style>
</head>
<body>
<header class="banner">
<div style="padding-top: 100px; padding-left: 200px; margin-bottom: -150px; width: 350px; height: auto;">
<a href="index.html">
<img src="images/geo_flake_b&w.png" alt="Manzano Analytics Icon" style="width: 100%; height: 100%; transition: transform 0.6s ease-in-out;"
class="hover-zoom">
</a>
</div>
</header>
</body>
<body class="is-preload" style="background-image: url(images/mountains_stars_2.jpeg); background-size: cover;">
<!-- Page Wrapper -->
<div id="page-wrapper">
<!-- Header -->
<header id="header">
<h1><a href="index.html">Manzano Analytics</a></h1>
<nav>
<a href="#menu">Menu</a>
</nav>
</header>
<!-- Menu -->
<nav id="menu">
<div class="inner" style="width: auto;">
<h2>Menu</h2>
<ul class="links">
<li><a href="index.html"><img src="images/geo_flake_b&w.png" style="width: 20px; height: auto; vertical-align: middle;"> Home</a></li>
<li><a href="resume.html" target="_blank" class="icon solid fa-gem"> Resume</a></li>
<li><a href="Women_Empowerment.html">The Economy of Women's Rights</a></li>
<li><a href="Rockbuster.html">RockBuster and XGBoost</a></li>
<li><a href="Influenza.html">Influenza Staffing Strategy</a></li>
<li><a href="Beecher.html">Archaelogical Analysis</a></li>
<li><a href="Instacart.html">Instacart Customer Analysis</a></li>
<li><a href="Gameco.html">GameCo Marketing Strategy</a></li>
<li><a href="Euro_Tuition.html">European Education Costs</a></li>
<li><a href="Pig_E_Bank.html">Bank Churn Modeling</a></li>
<li><a href="index.html#footer"><img src="images/geo_flake_double.png" style="width: 20px; height: auto; vertical-align: middle;"> Contact</a></li>
</ul>
<a href="#" class="close">Close</a>
</div>
</nav>
<!-- Wrapper -->
<section id="wrapper">
<header>
<div class="content2">
<h2>Bank Churn Modeling</h2>
<p>A mock bank dataset to model customer loyalty.</p>
</div>
<div class="image-container" style="position: absolute; right: 100px; top: 260px;">
<img src="images/Pig_intro.png" style="width: 380px; height: auto; border-width: 5px;">
</div>
</header>
<!-- Content -->
<div class="wrapper">
<div class="inner">
<h3 class="major">Modeling Bank Data</h3>
<p>The goal is to find quantifiable reasons that increase bank customers leaving. Immediately, modeling with
XGBoost or another machine learning model becomes a useful direction, provided model interpretability
is adequate. This is a perfect job for SHAP values. With a few simple lines of code, we've trained a model,
measured the accuracy of the model, and visualized the importance of each variable in the model's predictive
capability.
</p>
<img src="images/Pig_Python_Shap.png" style="width: 850px; height: auto; display: block; margin-left: auto; margin-right: auto; margin-bottom: 50px;">
<h3 class="major">Model Interpretability</h3>
<p>A number of variables stick out as important, and when put through a correlational analysis in Tableau,
the top SHAP values are confirmed as having higher correlations than other variables.
</p>
<img src="images/Pig_Shap_.png" style="width: 1000px; height: auto; display: block; margin-left: auto; margin-right: auto; margin-bottom: 50px;">
<h3 class="major">Decision Tree Mapping</h3>
<p>While the boosting in XGBoost does happen in an ensemble of decision trees, the feature importance and correlation
analysis gives us a solid framework to build our own decision tree from scratch.
</p>
<img src="images/Pig_Flow_chart.png" style="width: 800px; height: auto; display: block; margin-left: auto; margin-right: auto; margin-bottom: 50px;">
<section id="banner" style="display: block; margin: 0 auto;">
<div class="inner" >
<div class="logo"><img src="images/geo_flake_b&w.png" style="width: 150px; height: auto; display: block; margin: 0 auto; padding-bottom: 50px;"></i></div>
<div class="no-background">
<h2 style="font-size: 30px; display: block; margin: 0 auto; text-align: center;">Manzano Analytics: Bank Churn Modeling</h2>
<p style="font-size: 30px; font-weight: 100; text-align: center;">Lessons Learned:</p>
</div>
<div class="content1">
<section>
<div class="row">
<div class="alternate">
<h4 style="font-size: 18px;">SHAP Values</h4>
<ul style="font-size: 20px;">
<li>SHapely Additive exPlanations (SHAP values) are an important and versatile tool in an analysts arsenal.
In this project, I needed to create a decision tree from scratch, and needed to decide on the most important
features and values of those features. SHAP values were employed to identify important features. Additionally,
Tableau was utilized for correlational analysis with scatterplots to confirm or refute the findings of the
XGBoost model and SHAP values. Consistent SHAP and scatterplot correlations offered a robust framework for
predicting customer churn.
</li>
</ul>
<h4 style="font-size: 18px;">Decision Tree Modeling</h4>
<ul style="font-size: 20px;">
<li>Once the important features and their relevant values were identified, the decision tree could be created. Age, Number of Products,
and Account Balance emerged as the most critical variables once identified. Creating the decision tree and devising strategies to
reduce customer churn became straightforward by targeting high-risk customers with promotional and customer service efforts.
</li>
</ul>
</div>
</section>
</div>
</div>
</section>
<section class="features">
<article>
<a href="Rockbuster.html" class="image"><img src="images/Rockbuster_intro 2.png" alt="" style="padding-top: 50px; padding-bottom: 40px;"/></a>
<h3 class="major">Rockbuster and XGBoost</h3>
<p>This cost-benefit analysis is in a context of a fictional, globally operating movie rental business on the brink of expansion. Its primary goal is to employ strategic decisions that result in an augmentation of overall profitability.
</p>
<a href="Rockbuster.html" class="special" style="padding-top: 50px;">Learn more</a>
</article>
<article>
<a href="Women_Empowerment.html" class="image"><img src="images/WEI_K_Clusters_countries.png" alt="" /></a>
<h3 class="major">Women's Empowerment and GDP</h3>
<p>This is a five-decade examination of the interconnection between the Global Women's Empowerment Index and GDP Per Capita. The central proposition suggests a mutually reinforcing association between economic well-being and Women's Rights. Nevertheless, our unwavering dedication is to adopt a rigorous, data-centric approach, unveiling noteworthy discoveries.
</p>
<a href="#" class="special">Learn more</a>
</article>
</section>
</div>
</div>
</section>
<!-- Footer -->
<section id="footer">
<div class="inner">
<h2 class="major">Get in touch</h2>
<p> A seasoned and self-made fund manager, averaging a 40% CAGR for six years, and demonstrating an expertise in applied data analytics..</p>
<form method="post" action="https://formspree.io/f/mayrjnpo">
<div class="fields">
<div class="field">
<label for="name">Name</label>
<input type="text" name="name" id="name" />
</div>
<div class="field">
<label for="email">Email</label>
<input type="email" name="email" id="email" />
</div>
<div class="field">
<label for="message">Message</label>
<textarea name="message" id="message" rows="4"></textarea>
</div>
</div>
<ul class="actions">
<li><input type="submit" value="Send Message" /></li>
</ul>
</form>
<ul class="contact">
<li class="icon brands fa-codepen">
<b>Manzano Analytics</b><br />
Aaron Manzano<br />
Washington, PA 15301
</li>
<li class="icon solid fa-phone">(724) 787-9773</li>
<li class="icon solid fa-envelope"><a href="mailto:aaronlmanzano@gmail.com">AaronLManzano@gmail.com</a></li>
<li class="icon solid fa-gem"><a href="resume.html" target="_blank">Resume</a></li>
<li class="icon brands fa-linkedin"><a href="https://www.linkedin.com/in/aaronmanzano" target="_blank">Linkedin</a></li>
<li class="icon brands fa-github"><a href="https://github.com/tendyzen" target="_blank">GitHub</a></li>
</ul>
<ul class="copyright">
<li>© Manzano Analytics Inc. All rights reserved.</li><li>Design: <a href="http://html5up.net">HTML5 UP</a></li>
</ul>
</div>
</section>
</div>
<!-- Scripts -->
<script src="assets/js/jquery.min.js"></script>
<script src="assets/js/jquery.scrollex.min.js"></script>
<script src="assets/js/browser.min.js"></script>
<script src="assets/js/breakpoints.min.js"></script>
<script src="assets/js/util.js"></script>
<script src="assets/js/main.js"></script>
</body>
</html>