OP makes highly dubious claims and does not pass a basic smell test:
ALL CLAIMS MADE ARE DUBIOUS: Implementing 130+ papers in 7 months is highly implausible. This would require:
- Insane productivity.
- Implausible access to pricing and news data resources (which are often not freely available).
- Expertise in machine learning, natural language processing, finance, and data science. OP had to implement financial, time-series and linguistic feature engineering pipelines, as well infer the architecture and hyper-parameters used in all papers AND train all these models. ALL WITHIN 7 MONTHS. All while previously being a trader professionally i.e. not likely an expert in many of these fields.
- OP also claims he "web scraped" all the data which is highly unlikely as price datasets are often sold for a pretty penny and not publicly available in the detail described in several of these papers.
- Down the thread, OP says he does not know what a "meta analysis" study is al the while being capable of implementing 130+ papers. So someone who is an expert in ML, statistics, data science and finance does not know one of the most basic types of scientific study. All the while essentially engaging in a meta analysis study.
- OP describes himself as "a trader at a Tier 1 US bank" to lend credibility to his post: in itself that description is ridiculous and sounds like a naive attempt at instilling authority.
- When others encourage OP to publish results, he answers evasively: "probably a bit deep for a public forum but I was kinda glad to see the back of that work. It was an awesome learning experience but it's pretty soul destroying experimenting with tonnes of stuff that just doesn't work."
EVIDENCE PROVIDED: Non-existent.
All OP has to show for all this work is a hastily written Reddit post with dubious claims. There is no proof of the work done whatsoever, no code samples, not even result tables or graphs. The discussion of basic results are often made criticisms of this line of research.
MOTIVATION:
At the end OP shills his cryptotrading bot.
This post was likely all just purely made-up to market his cryptotrading bot service. OP uses some common criticisms of market prediction research to garner authority as a wizz-kid to attract people to his crypto scam.
What's worse many on HN and Reddit seem to gobble it up naively. Seemingly because OP is critical of something that is popular to criticize.
ALL CLAIMS MADE ARE DUBIOUS: Implementing 130+ papers in 7 months is highly implausible. This would require:
- Insane productivity.
- Implausible access to pricing and news data resources (which are often not freely available).
- Expertise in machine learning, natural language processing, finance, and data science. OP had to implement financial, time-series and linguistic feature engineering pipelines, as well infer the architecture and hyper-parameters used in all papers AND train all these models. ALL WITHIN 7 MONTHS. All while previously being a trader professionally i.e. not likely an expert in many of these fields.
- OP also claims he "web scraped" all the data which is highly unlikely as price datasets are often sold for a pretty penny and not publicly available in the detail described in several of these papers.
- Down the thread, OP says he does not know what a "meta analysis" study is al the while being capable of implementing 130+ papers. So someone who is an expert in ML, statistics, data science and finance does not know one of the most basic types of scientific study. All the while essentially engaging in a meta analysis study.
- OP describes himself as "a trader at a Tier 1 US bank" to lend credibility to his post: in itself that description is ridiculous and sounds like a naive attempt at instilling authority.
- When others encourage OP to publish results, he answers evasively: "probably a bit deep for a public forum but I was kinda glad to see the back of that work. It was an awesome learning experience but it's pretty soul destroying experimenting with tonnes of stuff that just doesn't work."
EVIDENCE PROVIDED: Non-existent.
All OP has to show for all this work is a hastily written Reddit post with dubious claims. There is no proof of the work done whatsoever, no code samples, not even result tables or graphs. The discussion of basic results are often made criticisms of this line of research.
MOTIVATION:
At the end OP shills his cryptotrading bot. This post was likely all just purely made-up to market his cryptotrading bot service. OP uses some common criticisms of market prediction research to garner authority as a wizz-kid to attract people to his crypto scam.
What's worse many on HN and Reddit seem to gobble it up naively. Seemingly because OP is critical of something that is popular to criticize.