Skip to content
/ ICML2025ML Public

Official dataset for the ICML 2025 New-In-ML Workshop paper “Event-Aware Sentiment Factors from Financial Tweets”, developed in collaboration with F Lab.

2 stars 0 forks Branches Tags Activity
Star
Notifications You must be signed in to change notification settings

yw562/ICML2025ML

BranchesTags

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
20172018
20172018
 
 
__pycache__
__pycache__
 
 
alpha_factor_extension
alpha_factor_extension
 
 
data
data
 
 
figs
figs
 
 
grid_results
grid_results
 
 
models
models
 
 
output
output
 
 
.DS_Store
.DS_Store
 
 
04_quantile_returns.png
04_quantile_returns.png
 
 
Day3BacktestPipelinenew.py
Day3BacktestPipelinenew.py
 
 
Day3enhanced backtestnew.py
Day3enhanced backtestnew.py
 
 
Day3enhancednew.py
Day3enhancednew.py
 
 
Latex_multihorizon.py
Latex_multihorizon.py
 
 
README.md
README.md
 
 
alpha_radar.png
alpha_radar.png
 
 
alpha_radar_analysis.py
alpha_radar_analysis.py
 
 
alpha_radar_analysis_significance.py
alpha_radar_analysis_significance.py
 
 
build_prices_from_cleaned.py
build_prices_from_cleaned.py
 
 
cleaned_aligned_data.csv
cleaned_aligned_data.csv
 
 
compare_cumulative_return_full_vs_truncated.png
compare_cumulative_return_full_vs_truncated.png
 
 
compare_max_drawdown_full_vs_truncated.png
compare_max_drawdown_full_vs_truncated.png
 
 
compare_sharpe_ratio_full_vs_truncated.png
compare_sharpe_ratio_full_vs_truncated.png
 
 
compare_topN_returns.png
compare_topN_returns.png
 
 
compare_topN_returns_filtered_truncated_day9.png
compare_topN_returns_filtered_truncated_day9.png
 
 
compare_topN_returns_truncated20171231.png
compare_topN_returns_truncated20171231.png
 
 
cumulative_return_1D.png
cumulative_return_1D.png
 
 
cumulative_return_2D.png
cumulative_return_2D.png
 
 
cumulative_return_3D.png
cumulative_return_3D.png
 
 
cumulative_return_7D.png
cumulative_return_7D.png
 
 
daily_ic_1_DAY_RETURN.png
daily_ic_1_DAY_RETURN.png
 
 
daily_ic_2_DAY_RETURN.png
daily_ic_2_DAY_RETURN.png
 
 
daily_ic_3_DAY_RETURN.png
daily_ic_3_DAY_RETURN.png
 
 
daily_ic_7_DAY_RETURN.png
daily_ic_7_DAY_RETURN.png
 
 
daily_winrate_1_DAY_RETURN.png
daily_winrate_1_DAY_RETURN.png
 
 
daily_winrate_2_DAY_RETURN.png
daily_winrate_2_DAY_RETURN.png
 
 
daily_winrate_3_DAY_RETURN.png
daily_winrate_3_DAY_RETURN.png
 
 
daily_winrate_7_DAY_RETURN.png
daily_winrate_7_DAY_RETURN.png
 
 
day10_advanced_strategy_analysis.py
day10_advanced_strategy_analysis.py
 
 
day11_ic_and_winrate.py
day11_ic_and_winrate.py
 
 
day12_t0_strategy_summary (1).py
day12_t0_strategy_summary (1).py
 
 
day13_gemini_labeling_google_ai_studio.py
day13_gemini_labeling_google_ai_studio.py
 
 
day13_huggingface_labeling_api.py
day13_huggingface_labeling_api.py
 
 
day13_prepare_llm_emotion_labeling.py
day13_prepare_llm_emotion_labeling.py
 
 
day2_export_signals.py
day2_export_signals.py
 
 
day2_sestm_pipeline_delete_noise.py
day2_sestm_pipeline_delete_noise.py
 
 
day4_result.csv
day4_result.csv
 
 
day4_simple_backtest.py
day4_simple_backtest.py
 
 
day4_strategy_performance.png
day4_strategy_performance.png
 
 
day5_metrics_visuals.py
day5_metrics_visuals.py
 
 
day6_compare_topN.py
day6_compare_topN.py
 
 
day6_compare_topN_truncated.py
day6_compare_topN_truncated.py
 
 
day7_compare_full_vs_truncated.py
day7_compare_full_vs_truncated.py
 
 
day8_strategy_diagnostics.py
day8_strategy_diagnostics.py
 
 
day9_filtered_backtest (1).py
day9_filtered_backtest (1).py
 
 
diagnostics_log.txt
diagnostics_log.txt
 
 
drawdown_plot.png
drawdown_plot.png
 
 
emotion_factor_analysis.py
emotion_factor_analysis.py
 
 
filter_2017_cleaned_aligned_data.csv
filter_2017_cleaned_aligned_data.csv
 
 
filter_2017_cleaned_aligned_data.py
filter_2017_cleaned_aligned_data.py
 
 
ic_across_horizon.png
ic_across_horizon.png
 
 
ic_summary.csv
ic_summary.csv
 
 
latex_compare_topn_sensitivity_plot.py
latex_compare_topn_sensitivity_plot.py
 
 
latex_strategy_combined_topN.py
latex_strategy_combined_topN.py
 
 
llm_emotion_type_labeling_samples.csv
llm_emotion_type_labeling_samples.csv
 
 
llm_emotion_type_labeling_samples_labeled_gemini.csv
llm_emotion_type_labeling_samples_labeled_gemini.csv
 
 
merged_latex_table.tex
merged_latex_table.tex
 
 
merged_latex_table2.tex
merged_latex_table2.tex
 
 
metrics_by_label.csv
metrics_by_label.csv
 
 
monthly_sharpe.png
monthly_sharpe.png
 
 
monthly_sharpe_table.csv
monthly_sharpe_table.csv
 
 
prices.csv
prices.csv
 
 
prices.parquet
prices.parquet
 
 
sharpe_across_horizon.png
sharpe_across_horizon.png
 
 
signal_count_timeseries.png
signal_count_timeseries.png
 
 
signal_distribution.png
signal_distribution.png
 
 
signal_strength_over_time.png
signal_strength_over_time.png
 
 
signals.parquet
signals.parquet
 
 
signals2017.parquet
signals2017.parquet
 
 
summary_metrics.csv
summary_metrics.csv
 
 
t0_strategy_summary_extended.csv
t0_strategy_summary_extended.csv
 
 
tone quantile return.png
tone quantile return.png
 
 
tone quantile return.py
tone quantile return.py
 
 
topN_sensitivity_plot.png
topN_sensitivity_plot.png
 
 
topN_stock_frequency.png
topN_stock_frequency.png
 
 
topN_strategy_summary.csv
topN_strategy_summary.csv
 
 
topN_strategy_summary_filtered_truncated_day9.csv
topN_strategy_summary_filtered_truncated_day9.csv
 
 
topN_strategy_summary_truncated20171231.csv
topN_strategy_summary_truncated20171231.csv
 
 
tweet_level_preds.csv
tweet_level_preds.csv
 
 
tweet_level_preds2017.csv
tweet_level_preds2017.csv
 
 
winrate_across_horizon.png
winrate_across_horizon.png
 
 
winrate_summary.csv
winrate_summary.csv
 
 
worst_stocks_post20171231.png
worst_stocks_post20171231.png
 
 

Repository files navigation

This code was developed as part of my work for a paper submitted to an ICML workshop. It uses only publicly available datasets, and all the code was entirely written by Chiyao and I. The project is open-sourced. I am affiliated with the University of Cambridge and Flab.

About

Official dataset for the ICML 2025 New-In-ML Workshop paper “Event-Aware Sentiment Factors from Financial Tweets”, developed in collaboration with F Lab.

Resources

Readme
Activity

Stars

2 stars

Watchers

0 watching

Forks

0 forks
Report repository

Releases

No releases published

Packages

No packages published

Languages