Giter Site home page Giter Site logo

kaggle-taxi-ii's Introduction

kaggle-taxi-trip-time-prediction

Winning solution for the Taxi-Trip Time Prediction Challenge on Kaggle (https://www.kaggle.com/c/pkdd-15-taxi-trip-time-prediction-ii).

Documentation about the method is available in doc/Taxi_II_Winning_Solution.pdf. Information on how to generate the solution file can be found below.

Generating the solution

Install the dependencies

I used python 2.7.9 with its popular scientific modules, in particular

Download the code

To download the code, run:

git clone git://github.com/hochthom/kaggle-taxi-ii.git

Download the training data

Download the data files from Kaggle. Place and extract the files into the data directory.

Create the training and test sets

Run the following scripts in any order.

create_training_set_N1.py

Result: train_pp_N1.csv with size: 1658559 x 8, Duration: 679.9 sec *

create_training_set_N2.py

Result: train_pp_N2.csv with size: 1658559 x 10, Duration: 771.8 sec *

create_training_set_N3.py

Result: train_pp_N3.csv with size: 1658559 x 10, Duration: 819.9 sec *

create_training_set.py

Result: train_pp_RND.csv with size: 1928087 x 17, Duration: 911.5 sec *

create_training_set_Experts.py

Result: 319 Training Sets, with size 3 - 109915 x 17, Duration: 10135.2 sec *
train_pp_TST_0.csv - train_pp_TST_319.csv, no file for trip 125.

*Note: The listed running times were achieved on a 12-core server with 24GByte of ram. The running time may be considerably longer if you use a machine with less ram!

Create the single model predictions

mk_submission.py

creates submissions for models N1, N2, N3, RND; Duration: 5175.5 sec *
Result: my_submission_xx.csv with xx in [N1,N2,N3,RND]

mk_submission_Experts.py

creates the expert model predictions, Duration: 2968.3 sec *
Result: predictions_TVT_experts.pkl

*Note: The listed running times were achieved on a 12-core server with 24GByte of ram. The running time may be considerably longer if you use a machine with less ram!

Create the final (blended) submissions

To generate blended prediction files from all the models, run:

blend_submissions.py

Result: final_submission_1.csv and final_submission_2.csv

Submit predictions

Submit the files on Kaggle to get it scored.

Submission 1 scored better in the private leaderboard than submission 2 (0.5092 vs. 0.5045), whereas for the public leaderboard it was the other way around (0.5253 vs. 0.5354).

kaggle-taxi-ii's People

Contributors

hochthom avatar

Stargazers

Nicole Gu avatar Camille Dunning avatar  avatar Yohanes Satria Nugroho avatar Marco Antônio Aburachid Tavares avatar Tanyapohn Patummasutr avatar  avatar B0Qi avatar benw avatar Hossam Ali avatar  avatar Md Hishamur Rahman avatar  avatar Rod Rivera avatar Regis Pires Magalhaes avatar 小猿猴GISer avatar CodeKitty avatar Rui Jia avatar  avatar  avatar  avatar Akizya avatar Chan, Yao-Chun(Eric) avatar Swapnil Phulse avatar YaGuang avatar  avatar Runze Li avatar  avatar Zuping Wu avatar  avatar  avatar Yingqian Mei avatar Vikram Dutt avatar Haswell Iris avatar Sefa Koç avatar  avatar  avatar pioneeryi avatar avinash avatar Carlos Bentes avatar MS avatar  avatar Dan Ofer avatar Bharat Gupta avatar  avatar Eyad Sibai avatar jiandong avatar theoky avatar

Watchers

James Cloos avatar  avatar Tony Gu avatar  avatar  avatar

kaggle-taxi-ii's Issues

No issue

I would like to be able to implement this on Quantum computer

Recommend Projects

  • React photo React

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

  • Vue.js photo Vue.js

    🖖 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. 📊📈🎉

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 ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.