Hull Tactical - Market Prediction

This guide outlines the steps to participate in the Kaggle competition for Hull Tactical - Market Prediction.
Hull Tactical - Market Prediction

1. Join Kaggle Competition

• Create a Kaggle account if you don't have one.
• Join the "Hull Tactical - Market Prediction" competition.
• Familiarize yourself with the rules, evaluation metric (Log Loss), and data format.

2. Notebook Setup

• Create a new notebook in Kaggle or your preferred environment.
• Add the competition dataset (/kaggle/input/hull-tactical-market-prediction/train.csv) to your notebook.

3. Environment Settings

• LGBM: Set Accelerator to "None". (But, it doesn't matter if "GPU T4 x2".)
• LSTM: Set Accelerator to "GPU T4 x2".

4. Run Notebook

• Execute the cells in your notebook sequentially to train models and generate predictions.
• Ensure your final submission. In this Kaggle competition, you don't submit a CSV file. Instead, you need to create and submit a prediction function as shown below.

import polars as pl
import kaggle_evaluation.default_inference_server

def predict(test: pl.DataFrame):
    # your prediction code
    return 0

inference_server = kaggle_evaluation.default_inference_server.DefaultInferenceServer(predict)