Inspired by Learnings from Kaggle’s Forecasting Competitions by Casper Solheim Bojer & Jens Peder Meldgaard in 2020, I surveyed the top 3 solutions in the past kaggle time series competitions since 2014 to 2023.
Table of Contents
- List of competitions
- Top 3 most voted EDAs
- 1. Walmart Recruiting - Store Sales Forecasting
- 2. Walmart Recruiting II: Sales in Stormy Weather
- 3. Rossmann Store Sales
- 4. Web Traffic Time Series Forecasting
- 5. TalkingData AdTracking Fraud Detection Challenge
- 6. Corporación Favorita Grocery Sales Forecasting
- 7. Recruit Restaurant Visitor Forecasting
- 8. Google Analytics Customer Revenue Prediction
- 9. LANL Earthquake Prediction
- 10. Two Sigma: Using News to Predict Stock Movements
- 11. ASHRAE - Great Energy Predictor III
- 12. University of Liverpool - Ion Switching
- 13. M5 Forecasting - Accuracy
- 14. Jane Street Market Prediction
- 15. Acea Smart Water Analytics
- 16. Google Brain - Ventilator Pressure Prediction
- 17. Optiver Realized Volatility Prediction
- 18. G-Research Crypto Forecasting
- 19. Ubiquant Market Prediction
- 20. American Express - Default Prediction
- 21. GoDaddy - Microbusiness Density Forecasting
- Top 3 solutions
- 1. Walmart Recruiting - Store Sales Forecasting
- 2. Walmart Recruiting II: Sales in Stormy Weather
- 3. Rossmann Store Sales
- 4. Web Traffic Time Series Forecasting
- 5. TalkingData AdTracking Fraud Detection Challenge
- 6. Corporación Favorita Grocery Sales Forecasting
- 7. Recruit Restaurant Visitor Forecasting
- 8. Google Analytics Customer Revenue Prediction
- 9. LANL Earthquake Prediction
- 10. Two Sigma: Using News to Predict Stock Movements
- 11. ASHRAE - Great Energy Predictor III
- 12. University of Liverpool - Ion Switching
- 13. M5 Forecasting - Accuracy
- 14. Jane Street Market Prediction
- 15. Acea Smart Water Analytics
- 16. Google Brain - Ventilator Pressure Prediction
- 17. Optiver Realized Volatility Prediction
- 18. G-Research Crypto Forecasting
- 19. Ubiquant Market Prediction
- 20. American Express - Default Prediction
- 21. GoDaddy - Microbusiness Density Forecasting
To learn the characteristic of data given in each competition, EDA is one of the best way.
So top 3 most voted EDAs are listed.
- EDA and Store Sales Predictions using XGB
- Walmart prediction - (1) EDA with time and space
- Wallmart Sales - EDA - feat eng [Future Update]
NA
- Time Series Analysis and Forecasts with Prophet
- EDA and forecasting with RFRegressor_FINAL_UPDATED
- How Does New Competition Affect Sales?
- Wiki Traffic Forecast Exploration - WTF EDA
- Web Traffic Time Series Forecasting (EDA)
- Wikipedia Web traffic EDA
- TalkingData EDA plus time patterns
- TalkingData EDA and Class Imbalance
- TalkingData: EDA to Model Evaluation | LB: 0.9683
- Shopping for Insights - Favorita EDA
- Memory optimization and EDA on entire dataset
- Grocery EDA Dirty XGBoost, Arima,ETS,Prophet
- R EDA for GStore + GLM + KERAS + XGB
- Google Analytics EDA + LightGBM + Screenshots
- A Very Extensive GStore Exploratory Analysis
- Earthquakes FE. More features and samples
- LANL Earthquake EDA and Prediction
- Masters Final Project: EDA
- EDA, feature engineering and everything
- 👨🔬 Bird Eye 👀 view of Two Sigma + NN Approach
- Simple EDA - Two Sigma
- Back to (predict) the future - Interactive M5 EDA
- M5 Competition : EDA + Models 📈
- Time Series Forecasting-EDA, FE & Modelling📈
- Jane Street: EDA of day 0 and feature importance
- Jane_street_Extensive_EDA & PCA starter 📊⚡
- EDA / A Quant's Prespective
- Acea Smart Water: Full EDA & Prediction
- EDA: Quenching the Thirst for Insights
- Quick EDA | Reporting & Data Understanding
- Ventilator Pressure Prediction: EDA, FE and models
- 🔥EDA +FE+TabNet 🧠🧠[Weights and Biases]
- Ventilator Pressure: EDA and simple submission
- Optiver Realized: EDA for starter(English version)
- Optiver Realized Volatility Prediction - EDA
- Optiver; EDA XGBoost starter(日本語,Japanese)
- 📊 G-Research Plots + EDA 📊
- To The Moon 🚀 [G-Research Crypto Forecasting EDA]
- 📈📊[G-crypto] Interactive Dashboard + Indicators
- AMEX EDA which makes sense ⭐️⭐️⭐️⭐️⭐️
- AMEX Default Prediction EDA & LGBM Baseline
- American Express EDA
| Pos | Methods | FE | Ensemble | Split | Code | Discussion |
|---|---|---|---|---|---|---|
| 1 | 💻 | 🔊 | ||||
| 2 | NA | 🔊 | ||||
| 3 | 💻 | 🔊 |
| Pos | Methods | FE | Ensemble | Split | Code | Discussion |
|---|---|---|---|---|---|---|
| 1 | Lasso | - | - | 💻 | 🔊 | |
| 2 | - | - | - | NA | NA | |
| 3 | - | - | - | NA | NA |
| Pos | Methods | FE | Ensemble | Split | Code | Discussion |
|---|---|---|---|---|---|---|
| 1 | NA | 🔊 | ||||
| 2 | NA | NA | ||||
| 3 | 💻 | 🔊 |
| Pos | Methods | FE | Ensemble | Split | Code | Discussion |
|---|---|---|---|---|---|---|
| 1 | 💻 | 🔊 | ||||
| 2 | 💻 | 🔊 | ||||
| 3 | 💻 | 🔊 |
| Pos | Methods | FE | Ensemble | Split | Code | Discussion |
|---|---|---|---|---|---|---|
| 1 | NA | 🔊 | ||||
| 2 | NA | 🔊 | ||||
| 3 | NA | 🔊 |
| Pos | Methods | FE | Ensemble | Split | Code | Discussion |
|---|---|---|---|---|---|---|
| 1 | 💻 💻 |
🔊 | ||||
| 2 | NA | 🔊 | ||||
| 3 | NA | 🔊 |
| Pos | Methods | FE | Ensemble | Split | Code | Discussion |
|---|---|---|---|---|---|---|
| 1 | LightGBM | - | - | 💻 💻 |
NA | |
| 2 | - | - | - | NA | NA | |
| 3 | - | - | - | NA | NA |
| Pos | Methods | FE | Ensemble | Split | Code | Discussion |
|---|---|---|---|---|---|---|
| 1 | 💻 | 🔊 | ||||
| 2 | NA | 🔊 | ||||
| 3 | - | - | - | NA | NA |
| Pos | Methods | FE | Ensemble | Split | Code | Discussion |
|---|---|---|---|---|---|---|
| 1 | 💻 💻 |
🔊 | ||||
| 2 | NA | 🔊 | ||||
| 3 | NA | 🔊 |
NA
| Pos | Methods | FE | Ensemble | Split | Code | Discussion |
|---|---|---|---|---|---|---|
| 1 | CatBoost LightGBM MLP |
NA | 🔊 | |||
| 2 | XGBoost LightGBM Catboost Feed-forward Neural Network |
NA | 🔊 | |||
| 3 | CNN LightGBM Catboost |
NA | 🔊 |
| Pos | Methods | FE | Ensemble | Split | Code | Discussion |
|---|---|---|---|---|---|---|
| 1 | NA | 🔊 | ||||
| 2 | 💻 | 🔊 🔊 |
||||
| 3 | NA | 🔊 🔊 |
| Pos | Methods | FE | Ensemble | Split | Code | Discussion |
|---|---|---|---|---|---|---|
| 1 | LightGBM | NA | 💻 | |||
| 2 | LightGBM | NA | 💻 | |||
| 3 | DeepAR | NA | 💻 |
| Pos | Methods | FE | Ensemble | Split | Code | Discussion |
|---|---|---|---|---|---|---|
| 1 | XGBoost NN |
💻 | NA | |||
| 3 | 49 layers MLPs | No | 15 ensembles of NN | NA | 🔊 |
NA for Pos #2
NA
| Pos | Methods | FE | Ensemble | Split | Code | Discussion |
|---|---|---|---|---|---|---|
| 1 | LSTM Transformer |
single architecture | KFold | 💻 | 🔊 | |
| 2 | Stacked LSTM | ensembled by 7 models | KFold | NA | 🔊 | |
| 3 | Conv1d Stacked LSTM |
random seed average | Stratified K-Folds | NA | 🔊 |
| Pos | Methods | FE | Ensemble | Split | Code | Discussion |
|---|---|---|---|---|---|---|
| 1 | LightGBM MLP CNN |
equally weighd average | GroupKFold | 💻 | 🔊 | |
| 3 | LightGBM MLP TabNet |
equally weighd average | KFold | 💻 | 🔊 |
NA for Pos #2
| Pos | Methods | FE | Ensemble | Split | Code | Discussion |
|---|---|---|---|---|---|---|
| 1 | - | - | - | - | NA | NA |
| 2 | LightGBM | Single model | NA | 🔊 | ||
| 3 | LightGBM | Single model | 💻 💻 | 🔊 |
| Pos | Methods | FE | Ensemble | Split | Code | Discussion |
|---|---|---|---|---|---|---|
| 1 | LightGBM TABNET |
Average of (LGBM x 5 Folds) + (TABNET x 5 Folds) | PurgedGroupTimeSeries TimeSerieseSplit KFold |
NA | 🔊 | |
| 2 | LightGBM | - | Purged K-FOLD cross validation with embargo | NA | 🔊 | |
| 3 | 6 layers transformer | 5 seeds ensemble | - | NA | 🔊 |
| Pos | Methods | FE | Ensemble | Split | Code | Discussion |
|---|---|---|---|---|---|---|
| 1 | LightGBM GRU |
Ensembled by 4 models | 💻 | 🔊 | ||
| 2 | LGB/XGB/CTB NN |
NA | 🔊 | |||
| 3 | LGB/CTB | Ensembled by 3 models | NA | 🔊 |
Ongoing
| Pos | Methods | FE | Ensemble | Split | Code | Discussion |
|---|---|---|---|---|---|---|
| 1 | ||||||
| 2 | ||||||
| 3 |