"Quite ingenious and insightful, and whereas your thoughts, perception and intuition are in precisely the correction market analysis dimension and direction...and as a former NYSE Market Maker this is the Master Key to absolute precision and laser accurate trade targeting and timing of trade executions..."
"Great article, creative feature engineering. I was looking at something like this too a little while ago and this has motivated me to try it again a little differently."
"Your writing always brings a smile to my face. Thank you for sharing your positivity and wisdom through this delightful article. Keep spreading joy!"
"Thanks for writing this! Very interesting, and useful in so many other cases besides stocks!"
"You present tech so elegantly"
"Thank you for sharing this intriguing approach to embeddings-based clustering. Seems promising. I am also contemplating transitioning to Vertex AI in the near future."
@sl_mar
Master mariner, MSc Physics & BSc Math, Law Scholar. Seeking alpha in finance using AI. Navigating sea and markets.
"Really, you mentioned very well the pain of financial experts who predict the markets. There are not enough mathematical methods to completely describe the most recent condition of the market and of course, they can not exactly predict. A good human expert can make predictions with a probability of success maximum of 60-65%. Nobody can make 100% correct predictions. There are some statistically proven models like gap close trading which have around 85-93% of confidence but they require 2-3x reserves for risks. Once you control and limit the risk, you will never get something better than 60% with a success rate and it will not guarantee that you will be profitable. That’s a reason why all professionals work with other people's capital, and create hedge funds, funds, and structured products. They just delegate the most difficult problems about risks to other people and share only the profits. Risk for investors, profits are for traders.
AI capacities always look great to solve a few problems at one time, so I would try AI in this field, not to try to make super correct predictions, but teach AI to work with the risks and calculate strategy not every decision."