https://github.com/jeffersonlicet/ml-titles-classification
💪 Public Leaderboard Top 10% - 0.89546 #MeliDataChallenge MercadoLibre Data Challenge
https://github.com/jeffersonlicet/ml-titles-classification
keras machine-learning nlp tensorflow
Last synced: about 2 months ago
JSON representation
💪 Public Leaderboard Top 10% - 0.89546 #MeliDataChallenge MercadoLibre Data Challenge
- Host: GitHub
- URL: https://github.com/jeffersonlicet/ml-titles-classification
- Owner: jeffersonlicet
- Created: 2019-09-08T01:14:17.000Z (almost 7 years ago)
- Default Branch: master
- Last Pushed: 2019-10-14T19:00:10.000Z (over 6 years ago)
- Last Synced: 2025-02-01T17:28:39.041Z (over 1 year ago)
- Topics: keras, machine-learning, nlp, tensorflow
- Language: Python
- Homepage: http://jeffersonlicet.com
- Size: 427 KB
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
### Public Leaderboard Top 20 - 0.89546
### The challenge
The challenge was very interesting, classify articles using only its titles.
### The prize
First and second place will receive tickets to [KHIPU](https://khipu.ai/). From 3 to 5 place an Intel Movidius.
### The result
I managed to be in the top 20 with a score of 0.8954, there were more of 150 participants, the competition was hard and exciting. Of course, i learned a lot of things.
## Preprocessing
Run ml_normalization.py to generate
- categories.npy
- titles.npy
- labels_normal.npy
## Generate Dictionary
Run ml_generate_dictionary.py to generate
- dictionary.npy
## Transform tokens
Run ml_index_tokens to generate
- hashed.npy
## Training the model
Run ml_train.py to train the model using
- hashed.npy titles as x data
- labels.npy as y data
- dictionary.npy to get vocabulary
## Preprocess test data
Run ml_test_normalization.py to normalize test data and generate
- test_titles.npy
## Generate submission file
Run ml_classify.py to generate the submission file