65 points by quantativeeric 1 year ago flag hide 11 comments
fintech_fanatic 1 year ago next
Fascinating topic! I've been following machine learning algorithms and their use in predicting stock market trends lately. Looking forward to reading the discussion.
ml_engineer 1 year ago next
I think combining both historical pricing data and sentiment analysis datapoints can give quite an accurate prediction of stock trends. I worked on a simulation-based model using LSTM and GRU networks. Will share more thoughts.
ml_engineer 1 year ago next
Absolutely, happy to share my experience. Our simulation using LSTM and GRU RNNs combined recent pricing data with news, Twitter and Reddit feeds. With enough labeled data, our model could anticipate bullish/bearish trends around 65%-70% of the time.
ml_engineer 1 year ago next
@quant_trader: Sure, it's unrealistic to achieve 100% accuracy with these models due to random alpha movements, over-optimization, and public news affecting the market irrationally. ML can work as a part of your arsenal but shouldn't be the sole factor.
robotics_researcher 1 year ago next
@ml_engineer: In your experience, what kind of preprocessing was needed to achieve the best results?
quant_trader 1 year ago prev next
While I agree that there is potential with ML techniques, I'm a strong believer in fundamental analysis and I'm yet to be fully convinced by the performance of these predictive models. Can any researchers provide some insight?
coding_enthusiast 1 year ago prev next
Could you please share some examples/repos on GitHub to help beginners like me get started with ML algorithms for predicting stock market trends?
deeplearning_geek 1 year ago next
@coding_enthusiast: Here's an example tutorial about "Predicting Stock Prices with Recurrent Neural Networks" on Kaggle for your reference: <https://www.kaggle.com/gpanayiotou/predicting-stock-prices-with-rnn>.
coding_enthusiast 1 year ago next
@deeplearning_geek: This is amazing! Thank you. Now I have a much clearer picture of how to get started.
coding_enthusiast 1 year ago next
@deeplearning_geek @ml_engineer @fintech_fanatic: What are the top ML libraries best suited for working... (continues)
web_dev 1 year ago prev next
What would you say is the more challenging part of these ML projects – creating a convincing model or preprocessing the data for usage?