Demo Day Page Google Classifier
| Demo Day Page Google Classifier | |
|---|---|
| Project Information | |
| Project Title | Demo Day Page Google Classifier |
| Owner | Kyran Adams |
| Start Date | 2/5/2018 |
| Deadline | |
| Keywords | Accelerator, Demo Day, Google Result, Word2vec, Tensorflow |
| Primary Billing | |
| Notes | |
| Has project status | Active |
| Is dependent on | Accelerator Seed List (Data), Demo Day Page Parser |
| Copyright © 2016 edegan.com. All Rights Reserved. | |
Project
This is a tensorflow project that classifies webpages as either a demo day page or not, currently using logistic regression. The classifier itself should take the output of Peter's DemoDayHits.py program and output whether the page is a demo day page. It is trained on a file outputted by DemoDayHits.py and a hand-classified set of google results, some of which are demo day pages.
It may later take other inputs, such as the text of the page itself.
A demo day page is an advertisement page for a "demo day," which is a day that cohorts graduating from accelerators can pitch their ideas to investors. These demo days give us a good idea of when these cohorts graduated from their accelerator.
The random forest implementation doesn't work on windows, so it is located in the Z drive to be run from the linux box.
Located:
E:\McNair\Projects\Accelerators\Spring 2018\google_classifier\ Z:\demoday
Training data:
E:\McNair\Projects\Accelerators\Fall 2017\Demo Day URLs.xlsx
Possibly useful programs
Google bindings for python
E:\McNair\Projects\Accelerators\Spring 2017\Google_SiteSearch
PDF to text converter
E:\McNair\Projects\Accelerators\Fall 2017\Code+Final_Data\Utilities\PDF_Ripper
HTML to text converted
E:\McNair\Projects\Accelerators\Fall 2017\Code+Final_Data
Resources
- https://www.tensorflow.org/tutorials/word2vec
- https://machinelearnings.co/tensorflow-text-classification-615198df9231
- http://www.wildml.com/2015/12/implementing-a-cnn-for-text-classification-in-tensorflow/
- https://stats.stackexchange.com/questions/181/how-to-choose-the-number-of-hidden-layers-and-nodes-in-a-feedforward-neural-netw
