ankishb.github.io/temp.html at master · ankishb/ankishb.github.io · GitHub

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
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351

<!DOCTYPE html>
<html>
<head>
     <meta name="viewport" content="width=device-width, initial-scale=1">
    <link rel="stylesheet" href="https://maxcdn.bootstrapcdn.com/bootstrap/3.4.1/css/bootstrap.min.css">
    <script src="https://ajax.googleapis.com/ajax/libs/jquery/3.4.1/jquery.min.js"></script>
    <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.4.1/js/bootstrap.min.js"></script>

<style>
.box {
     position: relative;
     width: 420px;
     height: 580px;
     border: 1px solid blue;
     box-sizing: border-box;
}
.box-s {
     position: relative;
     width: 900px;
     height: 600px;
     border: 1px solid blue;
     box-sizing: border-box;
}

.box-text {
     /*margin-top: -20px;*/
     width: 420px;
     height: 170px;
     margin-left: 2px;
     margin-right: 2px;
}
.box-s-text {
     /*margin-top: -20px;*/
     width: 900px;
     height: 180px;
     margin-left: 10px;
     margin-right: 10px;
}

.box-footer{
     position: absolute;
     bottom: 10px;
     vertical-align: bottom;
     display: inline-block;
     margin-bottom: -10px;
     width: 418px;
     height: 60px;
     background-color: lightgrey;
}
.box-s-footer{
     text-align: center;
     position: absolute;
     font-size: 24px;
     bottom: 10px;
     vertical-align: bottom;
     display: inline-block;
     padding-top: 5px;
     margin-bottom: -10px;
     width: 898px;
     height: 60px;
     background-color: lightblue;
}
</style>
</head>
<body>


<div class="container">
     <div class="box-s">
          <center><img align="middle" width="800" height="240" src="assets/images1/data-science.jpg"></center>
          <div class="box-s-block">
               <h2 align="middle">ML Toolbox</h2>
               <p class="box-s-text">This repo contains various data science strategy and machine learning models to deal with structure as well as unstructured data. It contains module on feature-preprocessing, feature-engineering, machine-learning-models, etc. Some of these features are collected from the existed libraries such as scikit-learn, keras, gensim, h2o, bayesopt, xgboost, lightgbm, catboost, GraphX etc and others are implemented by me, on following the Research Paper and Data-Scientist advice (on kaggle). I have worked on feature engineering strategy a lot, which you can check in my repo.</p>
          </div>

          <div class="box-s-buttons">
               <a href="https://github.com/ankishb/ml-toolbox" target="_blank" class="btn btn-primary">Github Repo</a>
          </div>

          <div class="box-s-footer">
               <small class="box-muted">Python, Gensim, Scikit-learn, H2o, bayesopt, GraphX, Keras, Xgboost, Catboost, lightgbm</small>
          </div>
      </div>
</div>

<br>
<div class="container">
     <h2>My Data science Projects</h2>

<div class="table-responsive">
<table class="table">
<tbody>
<tr>
     <td>
          <div>
          <div class="box">
               <center><img align="middle" width="400" height="240" src="assets/images1/Object-detection.png"></center>
               <div class="box-block">
                    <h3 class="box-title" align="middle">Class Agnostic Object Detection</h3>
                    <p class="box-text">A single shot detector algorithm in addition to object conditional feature, to create a robust bounding box around object. Using this method, I secured a rank of 36/6733 in the competition at dare2complete platform, sponsored by Flipkart.</p>
               </div>

               <div class="card-buttons">
                    <a href="https://github.com/ankishb/ml-projects/tree/master/conditional-object-detection" target="_blank" class="btn btn-primary">Github Repo</a>
                    <a href="https://github.com/ankishb/ml-projects/tree/master/conditional-object-detection" target="_blank" class="btn btn-primary">Project Report</a>
               </div>


               <div class="box-footer">
                    <small class="box-muted">Python, Tensorflow,  ImageAug, YOLO, SSD</small>
               </div>
           </div>
          </div>
     </td>
     <td>
          <div>
          <div class="box">
               <center><img align="middle" width="400" height="240" src="assets/images1/text-review.png"></center>
               <div class="box-block">
                    <h3 class="box-title" align="middle">Text classification</h3>
                    <p class="box-text">An ensemble of classifiers such as SVM, Logistic Regression, Decision Tree, Attention Model and BERT to accurately categorize the Amazon Product Review.</p>
               </div>

               <div class="box-buttons">
                    <a href="https://github.com/ankishb/ml-projects/tree/master/amazon-ml" target="_blank" class="btn btn-primary">Project</a>
               </div>

               <div class="box-footer">
                    <small class="box-muted">Python, EDA, Scikit-Learn, Xgboost, Keras, Word-Embedding, BERT, Ensemble</small>
               </div>
           </div>
          </div>
     </td>
</tr>

<tr><td><br></td></tr>

<tr>
     <td>
          <div>
          <div class="box">
               <center><img align="middle" width="400" height="240" src="assets/images1/credit-risk.jpg"></center>
               <div class="box-block">
                    <h3 class="box-title" align="middle">HDFC Risk Prediction</h3>
                    <p class="box-text">In this project, the challege was to build a model for 2500 unknown predictors(features). During this project, I experimented on a lot of feature selection and feature engineering strategy, to deal with such kind of situation. </p>
               </div>

               <div class="box-buttons">
                    <a href="https://github.com/ankishb/ml-projects/tree/master/hdfc-ml" target="_blank" class="btn btn-primary">Project</a>
               </div>

               <div class="box-footer">
                    <small class="box-muted">Python, EDA, Scikit-Learn, statsmodels, Gradient-Boosting, H2o, Stack-Net</small>
               </div>
           </div>
          </div>
     </td>
     <td>
          <div>
          <div class="box">
               <center><img align="middle" width="400" height="240" src="assets/images1/social-network.jpg"></center>
               <div class="box-block">
                    <h3 class="box-title" align="middle">Hike Friend Recommendation</h3>
                    <p class="box-text">This is a link prediction challenge. Here challenge is to handle a very big dataset, which doesn't fit in memory. During this project, I worked on graph network based feature engineering, graph embedding, memory optimization etc and secured a rank of 32/5389 in ML-Hikeathon contest.</p>
               </div>

               <div class="box-buttons">
                    <a href="https://github.com/ankishb/ml-projects/tree/master/hike-friend-recommendation" target="_blank" class="btn btn-primary">Project</a>
               </div>

               <div class="box-footer">
                    <small class="box-muted">GraphX, Gradient-Boost, Keras, Word-Embedding, Subsemble</small>
               </div>
           </div>
          </div>
     </td>
</tr>

<tr><td><br></td></tr>

<tr>
     <td>
          <div>
          <div class="box">
               <center><img align="middle" width="400" height="240" src="assets/images1/image-classification.png"></center>
               <div class="box-block">
                    <h3 class="box-title" align="middle">Cifar-10 Classification</h3>
                    <p class="box-text">Developed a architecture using state of the art ResNet, with class conditional feature. This add more attention on the class specific feature using global as well as local features. It improve the ResNet model's score by 1.37% on Cifar-10 Dataset.</p>
               </div>

               <div class="box-buttons">
                    <a href="https://github.com/ankishb/ml-projects/tree/master/cifar-10-resnet" target="_blank" class="btn btn-primary">Project</a>
               </div>

               <div class="box-footer">
                    <small class="box-muted">Python, Tensorflow, Attention Mechanism, Visulization</small>
               </div>
           </div>
          </div>
     </td>
     <td>
          <div>
          <div class="box">
               <center><img align="middle" width="400" height="240" src="assets/images1/facenet.jpg"></center>
               <div class="box-block">
                    <h3 class="box-title" align="middle">Face Verification system</h3>
                    <p class="box-text">The objective is to build a student-attendence system which takes image as input and verify if the person is registered for the course or not. As dataset was very limited, I used network in network architecture along with hard-mining technique to successfully train a matching network model, to achieve 93% accuracy.</p>
               </div>

               <div class="box-buttons">
                    <a href="https://github.com/ankishb/ml-projects/tree/master/facenet" target="_blank" class="btn btn-primary">Project</a>
               </div>

               <div class="box-footer">
                    <small class="box-muted">Tensorflow, Keras, ImageAug</small>
               </div>
           </div>
          </div>
     </td>
</tr>

<tr><td><br></td></tr>

<tr>
     <td>
          <div>
          <div class="box">
               <center><img align="middle" width="400" height="240" src="assets/images1/segmentation1.png"></center>
               <div class="box-block">
                    <h3 class="box-title" align="middle">Segmentation</h3>
                    <p class="box-text">Implemented an U-Net architecture on blood cell Dataset and fully convolutional network on traffic-street dataset. Finally experimented with generative adverserial network for better generalization in the presence of limited dataset.</p>
               </div>

               <div class="box-buttons">
                    <a href="https://github.com/ankishb/ml-projects/tree/master/segmentation" target="_blank" class="btn btn-primary">Project</a>
               </div>

               <div class="box-footer">
                    <small class="box-muted">Python, Tensorflow, Keras, FCN, U-Net, GAN</small>
               </div>
           </div>
          </div>
     </td>
     <td>
          <div>
          <div class="box">
               <center><img align="middle" width="400" height="240" src="assets/images1/recommender-sytem.jpg"></center>
               <div class="box-block">
                    <h3 class="box-title" align="middle">Recommender System</h3>
                    <p class="box-text"></p>
               </div>

               <div class="box-buttons">
                    <a href="https://github.com/ankishb/ml-projects/tree/master/recommendation-system" target="_blank" class="btn btn-primary">Project</a>
               </div>

               <div class="box-footer">
                    <small class="box-muted">Python, Tensorflow, Word-Embedding, Graph, Gradient-Boosting, Ensemble</small>
               </div>
           </div>
          </div>
     </td>
</tr>

<tr><td><br></td></tr>

<tr>
     <td>
          <div>
          <div class="box">
               <center><img align="middle" width="400" height="240" src="assets/images1/stock-market.jpg"></center>
               <div class="box-block">
                    <h3 class="box-title" align="middle">Segmentation</h3>
                    <p class="box-text">Implemented an U-Net architecture on blood cell Dataset and fully convolutional network on traffic-street dataset. Finally experimented with generative adverserial network for better generalization in the presence of limited dataset.</p>
               </div>

               <div class="box-buttons">
                    <a href="https://github.com/ankishb/ml-projects/tree/master/segmentation" target="_blank" class="btn btn-primary">Project</a>
               </div>

               <div class="box-footer">
                    <small class="box-muted">Python, Tensorflow, Keras, FCN, U-Net, GAN</small>
               </div>
           </div>
          </div>
     </td>
     <td>
          <div>
          <div class="box">
               <center><img align="middle" width="400" height="240" src="assets/images1/text-mining.jpg"></center>
               <div class="box-block">
                    <h3 class="box-title" align="middle">Recommender System</h3>
                    <p class="box-text"></p>
               </div>

               <div class="box-buttons">
                    <a href="https://github.com/ankishb/ml-projects/tree/master/recommendation-system" target="_blank" class="btn btn-primary">Project</a>
               </div>

               <div class="box-footer">
                    <small class="box-muted">Python, Tensorflow, Word-Embedding, Graph, Gradient-Boosting, Ensemble</small>
               </div>
           </div>
          </div>
     </td>
</tr>
</tbody>
</table>

</div>


<br>
<h1>Small Fun Projects</h1>
<ul>
     <li class="list-group-item">
          <div class="d-flex w-100 justify-content-between">
               <h5>Behavioral Risk Factor Surveillance System 2013: Exploratory Data Analysis</h5>
               <a target="_blank" href="/portfolio/brfss2013_eda">Notebook</a>
          </div>
          <p>Analysis of the BRFSS-2013 data set using R, focusing on investigating the relationship between education and eating habits, sleep and mental health, and smoking, drinking and general health of a person.</p>
          <p class="list-group-skills">Skills: R, Descriptive Statistics, ggplot, dplyr</p>
     </li>

     <li class="list-group-item">
          <div class="d-flex w-100 justify-content-between">
               <h5>3-way Sentiment Analysis for Tweets</h5>
               <a target="_blank" href="/portfolio/sentiment_analysis_tweets">Notebook</a>
          </div>
          <p>3-way polarity (positive, negative, neutral) classification system for tweets, without using NLTK's sentiment analysis engine.</p>
          <p class="list-group-skills">Skills: Python, NLP, Scikit-learn</p>
     </li>
     <li class="list-group-item">
          <div class="d-flex w-100 justify-content-between">
               <h5>Inferential Statistics: Do men or women oppose sex education?</h5>
               <a target="_blank" href="/portfolio/gss_inferential">Notebook</a>
          </div>
          <p>Analysing the GSS (General Social Survey) dataset using R to infer if, in the year 2012, were men, of 18 years or above in the United States, more likely to oppose sex education in public schools than women.</p>
          <p class="list-group-skills">Skills: R, Hypothesis Testing, ggplot, dplyr</p>
     </li>
</ul>


     <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js"></script>

</body>
</html>