| <sub>[Stock-Prediction-Models](https://github.com/huseinzol05/Stock-Prediction-Models)</sub> | <sub>very good curated list of notebooks showing deep learning + reinforcement learning models. Also contain topics on outlier detections/overbought oversold study/monte carlo simulartions/sentiment analysis from text (text storage/parsing is not detailed but it mentioned using [BERT](https://github.com/google-research/bert))</sub> | <sub>12/18/17 10:49</sub> | <sub>1/5/21 10:31</sub> | <sub>3599.0</sub> | <sub>:heavy_check_mark:</sub> | <sub>:star:x5</sub> |
| <sub>[FinRL-Library](https://github.com/AI4Finance-LLC/FinRL-Library)</sub> | <sub>started by Columbia university engineering students and designed as an end to end deep reinforcement learning library for automated trading platform. Implementation of DQN DDQN DDPG etc using PyTorch and [gym](https://gym.openai.com/) use [pyfolio](https://github.com/quantopian/pyfolio) for showing backtesting stats. Big contributions on Proximal Policy Optimization (PPO) advantage actor critic (A2C) and Deep Deterministic Policy Gradient (DDPG) agents for trading</sub> | <sub>7/26/20 13:18</sub> | <sub>4/3/21 23:21</sub> | <sub>1807.0</sub> | <sub>:heavy_check_mark:</sub> | <sub>:star:x5</sub> |
| <sub>[RLTrader](https://github.com/notadamking/RLTrader)</sub> | <sub>predecessor to [tensortrade](https://github.com/tensortrade-org/tensortrade) uses open api [gym](https://gym.openai.com/) and neat way to render matplotlib plots in real time. Also explains LSTM/data stationarity/Bayesian optimization using [Optuna](https://github.com/optuna/optuna) etc.</sub> | <sub>4/27/19 18:35</sub> | <sub>10/17/19 16:25</sub> | <sub>1304.0</sub> | <sub>:heavy_check_mark:</sub> | <sub>:star:x5</sub> |
| <sub>[Personae](https://github.com/Ceruleanacg/Personae)</sub> | <sub>implementation of deep reinforcement learning and supervised learnings covering areas: deep deterministic policy gradient (DDPG) and DDQN etc. Data are being pulled from [rqalpha](https://github.com/ricequant/rqalpha) which is a python backtest engine and have a nice docker image to run training/testing</sub> | <sub>3/10/18 11:22</sub> | <sub>9/2/18 17:21</sub> | <sub>1142.0</sub> | <sub>:heavy_multiplication_x:</sub> | <sub>:star:x5</sub> |
| <sub>[awesome-deep-trading](https://github.com/cbailes/awesome-deep-trading)</sub> | <sub>curated list of papers/repos on topics like CNN/LSTM/GAN/Reinforcement Learning etc. Categorized as deep learning for now but there are other topics here. Manually maintained by cbailes</sub> | <sub>11/26/18 3:23</sub> | <sub>1/1/21 9:41</sub> | <sub>541.0</sub> | <sub>:heavy_check_mark:</sub> | <sub>:star:x4</sub> |
| <sub>[crypto-rl](https://github.com/sadighian/crypto-rl)</sub> | <sub>Retrieve limit order book level data from coinbase pro and bitfinex -> record in [arctic](https://github.com/man-group/arctic) timeseries database then implemented trend following strategies (market orders) and market making (limit orders). Uses reinforcement learning (DQN) [keras-rl](https://github.com/keras-rl/keras-rl) to create agents and uses [openai gym](https://gym.openai.com/) to implement POMDP (partially observable markov decision process)</sub> | <sub>6/21/18 1:06</sub> | <sub>11/5/20 11:08</sub> | <sub>345.0</sub> | <sub>:heavy_check_mark:</sub> | <sub>:star:x3</sub> |
| <sub>[Advanced-Deep-Trading](https://github.com/Rachnog/Advanced-Deep-Trading)</sub> | <sub>notebooks containing experiments based on Lopez de Prado book "Advances in financial machine learning". Mostly not deep learning related but rather sklearn regression models. Interesting libraries include [mlfinlab](https://github.com/hudson-and-thames/mlfinlab) for calculating return stats and [shap](https://github.com/slundberg/shap) for explaining models. Examlpe of shap can be which features are pushing the value up and and which features are pushing the value down. Also contain functions for calculating geometric brownian motion and jump diffusion functions. </sub> | <sub>2/16/19 21:18</sub> | <sub>11/29/20 20:12</sub> | <sub>319.0</sub> | <sub>:heavy_check_mark:</sub> | <sub>:star:x3</sub> |
| <sub>[BitcoinForecast](https://github.com/PiSimo/BitcoinForecast)</sub> | <sub>RNN model to predict short term price movement (in this case BTC for the next 9 minutes) [deepchart](https://pisimo.github.io/DeepChart/) is used to visualize the model </sub> | <sub>3/10/17 10:52</sub> | <sub>6/11/18 8:07</sub> | <sub>288.0</sub> | <sub>:heavy_multiplication_x:</sub> | <sub>:star:x3</sub> |
| <sub>[trading-bot](https://github.com/pskrunner14/trading-bot)</sub> | <sub>Implementation of deep reinforcement learning using Deep Q Network (DQN). Only supports single security at the moment. Idea is roughly based [here](https://keon.github.io/deep-q-learning/) and uses tensorflow/keras. Interesting helper python libraries used here are [tqdm](https://tqdm.github.io/) for console based progress bar and [altair](https://altair-viz.github.io/) for declarative visualization in python </sub> | <sub>8/13/18 10:44</sub> | <sub>1/23/20 4:41</sub> | <sub>286.0</sub> | <sub>:heavy_check_mark:</sub> | <sub>:star:x3</sub> |
| <sub>[DeepLearningInFinance](https://github.com/sonaam1234/DeepLearningInFinance)</sub> | <sub>Based on a [talk](https://towardsdatascience.com/deep-learning-in-finance-9e088cb17c03) Sonam Srivastava gave and there are two studies: 1. single timeseries return prediction using ARIMA/VAR/SVR/Deep Regression/CNN/LSTM 2. indexed portfolio construction using autoencoders i.e. replicate a index using handful of stocks. </sub> | <sub>8/21/17 16:00</sub> | <sub>8/21/17 17:23</sub> | <sub>266.0</sub> | <sub>:heavy_multiplication_x:</sub> | <sub>:star:x3</sub> |
| <sub>[Deep-Learning-Machine-Learning-Stock](https://github.com/LastAncientOne/Deep-Learning-Machine-Learning-Stock)</sub> | <sub>curated list of notebooks for machine learning models. Start with very simple linear models to more advanced reinforcement learning type of models. Problem with this repo is that the library version numbers may be changing over time and there's no specific way to track and upgrade</sub> | <sub>9/29/18 23:38</sub> | <sub>3/18/21 3:16</sub> | <sub>264.0</sub> | <sub>:heavy_check_mark:</sub> | <sub>:star:x3</sub> |
| <sub>[deep-RL-trading](https://github.com/golsun/deep-RL-trading)</sub> | <sub>trading game comparing RNN vs CNN vs MLP based on [paper](https://arxiv.org/abs/1803.03916)</sub> | <sub>2/25/18 17:41</sub> | <sub>12/1/20 22:06</sub> | <sub>233.0</sub> | <sub>:heavy_check_mark:</sub> | <sub>:star:x3</sub> |
| <sub>[Deep-Reinforcement-Stock-Trading](https://github.com/Albert-Z-Guo/Deep-Reinforcement-Stock-Trading)</sub> | <sub>inspired by Q-trader a deep reinforcement learning repo for trading. Only 3 actions allowed (buy/hold/sell) and no transaction cost is implemented yet. Uses [empyrical](https://github.com/quantopian/empyrical) for portfolio stats</sub> | <sub>5/19/19 22:20</sub> | <sub>9/27/20 19:22</sub> | <sub>141.0</sub> | <sub>:heavy_check_mark:</sub> | <sub>:star:x3</sub> |
| <sub>[Deep-Reinforcement-Learning-in-Trading](https://github.com/saeed349/Deep-Reinforcement-Learning-in-Trading)</sub> | <sub>Deep reinforcement learning for trading leveraging [openai gym](https://gym.openai.com/) framework. Keras implementation of DQN DDQN (double deep Q network) and DDDQN (dueling double dqn) trained/tested on s&p 500 daily data from 2013 to 2018. approach is described in an article [here](https://www.linkedin.com/pulse/deep-reinforcement-learning-trading-saeed-rahman)</sub> | <sub>5/11/18 0:52</sub> | <sub>10/26/19 14:22</sub> | <sub>137.0</sub> | <sub>:heavy_check_mark:</sub> | <sub>:star:x3</sub> |
| <sub>[AutomatedStockTrading-DeepQ-Learning](https://github.com/sachink2010/AutomatedStockTrading-DeepQ-Learning)</sub> | <sub>cornerstone project repo for Udacity nanodegree program [Become a machine learning engineer](https://www.udacity.com/course/machine-learning-engineer-nanodegree--nd009t) and focus on trading using deep q learning. Good explanation on design choices in the report</sub> | <sub>2/23/19 12:01</sub> | <sub>2/25/20 18:16</sub> | <sub>134.0</sub> | <sub>:heavy_check_mark:</sub> | <sub>:star:x3</sub> |
