Giter VIP home page Giter VIP logo

classifying-crisis-reports-dsc's Introduction

Growing Instability: Classifying Crisis Reports

This is a top 10 solution to the "Growing Instability: Classifying Crisis Reports" challenge from datasciencechallenge.org

The solution outline:

  • Use only the data from 2005 onwards for training
  • Discard the labels with less than 30 examples
  • Vectorize the texts, apply TF-IDF weights
  • For each label train an SVM
  • Use out-of-fold predictions for train for selecting the best threshold
  • For topics present in test and not present in train do a simple keyword matching

classifying-crisis-reports-dsc's People

Contributors

alexeygrigorev avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar  avatar

classifying-crisis-reports-dsc's Issues

calculation is too long more than 12 hours and still running

how long calculation should be?
I use lap top with 16 GB RAM , simulate during 12 hours and still running
it uses 10 GB RAM
how it should be, how long , what output looks like? ,
if I am doing something wrong?
Spasibo
Sander
may we communicate by email?
it is what I see on screen for 12 hours

0%| | 0/30 [00:00<?, ?it/s]
3%|3 | 1/30 [00:02<01:13, 2.54s/it]
7%|6 | 2/30 [00:05<01:14, 2.68s/it]
10%|# | 3/30 [00:09<01:18, 2.92s/it]
13%|#3 | 4/30 [00:12<01:21, 3.13s/it]
17%|#6 | 5/30 [00:16<01:22, 3.30s/it]
20%|## | 6/30 [00:19<01:21, 3.40s/it]
23%|##3 | 7/30 [00:23<01:21, 3.53s/it]
27%|##6 | 8/30 [00:27<01:19, 3.60s/it]
30%|### | 9/30 [00:31<01:19, 3.79s/it]
33%|###3 | 10/30 [00:35<01:17, 3.86s/it]
37%|###6 | 11/30 [00:40<01:15, 3.99s/it]
40%|#### | 12/30 [00:44<01:13, 4.06s/it]
43%|####3 | 13/30 [00:48<01:10, 4.13s/it]
47%|####6 | 14/30 [00:53<01:07, 4.22s/it]
50%|##### | 15/30 [00:58<01:07, 4.53s/it]
53%|#####3 | 16/30 [01:03<01:06, 4.73s/it]
57%|#####6 | 17/30 [01:09<01:05, 5.01s/it]
60%|###### | 18/30 [01:14<01:02, 5.20s/it]
63%|######3 | 19/30 [01:20<00:59, 5.38s/it]
67%|######6 | 20/30 [01:24<00:49, 4.96s/it]
70%|####### | 21/30 [01:28<00:42, 4.75s/it]
73%|#######3 | 22/30 [01:34<00:39, 4.89s/it]
77%|#######6 | 23/30 [01:39<00:35, 5.02s/it]
80%|######## | 24/30 [01:44<00:30, 5.15s/it]
83%|########3 | 25/30 [01:50<00:26, 5.24s/it]
87%|########6 | 26/30 [01:55<00:21, 5.27s/it]
90%|######### | 27/30 [02:00<00:15, 5.21s/it]
93%|#########3| 28/30 [02:06<00:10, 5.29s/it]
97%|#########6| 29/30 [02:11<00:05, 5.39s/it]
100%|##########| 30/30 [02:17<00:00, 5.35s/it]

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.