Corgipower12

11 ms·

1.
by Corgipower12 1y ago
That is so interesting! I wish you would write something up about it :) I guess that is why big companies mine social media data
2.
by Corgipower12 1y ago
YouTube video about the financial influencer research: https://youtu.be/A8TD6Oage4E?si=m3yuqIO0pvivSRa2
3.
by Corgipower12 1y ago
Yep! Markets are theoretically efficient.
4.
by Corgipower12 1y ago
Studies have been done that show that. Of course, if it is an old study, results need to be run on new data.
5.
by Corgipower12 1y ago
Your thought is that monetization often depends more on visibility than performance and that getting views on any platform these days also requires marketing? Fair point.
6.
by Corgipower12 1y ago
Like anything, I am sure there is good work out there. The keys is properly identifying which are good and which are bad.
7.
by Corgipower12 1y ago
Influencers often market things and make money off things other than being good at [name a skill or thing they are influencing]. For finance previous work has shown that financial influencers are are worse advice givers are actually the mor
8.
by Corgipower12 1y ago
I know that is probably sarcasm, but betting against things people are sharing might be interesting.
9.
by Corgipower12 1y ago
My 2 cents TLDR: seize opportunity but based on sound analysis and caution, not blind optimism. I think they are saying something of the effect of: 1. The short trade (The Big Short or similar trades during the housing bubble) happened duri
10.
by Corgipower12 1y ago
I wonder if it is because we hear about people who are successful betting against the herd.
11.
by Corgipower12 1y ago
I haven't benchmarked inverse Cramer recently, but last I heard not too well: https://www.reddit.com/r/wallstreetbets/comments/187612o/inv... Anyone know how it is doing recently?
12.
by Corgipower12 1y ago
There is a lot of psychology in the reasons why.
13.
by Corgipower12 1y ago
Investing in the QQQ or S&P 500 can often be a better idea than doing literally nothing.
14.
Betting against YouTube Financial Influencers beat the S&P 500 (risky though)?(papers.ssrn.com)
23 points by Corgipower12 1y ago|31 comments
15.
by Corgipower12 1y ago
We analyzed hundreds of stock recommendation videos from finance YouTubers (aka finfluencers) and backtested the results. Turns out, doing the opposite of what they say—literally inverting the advice—beat the S&P 500 by over +6.8% in an
16.
Facebook’s own PR shows Frances Haugen is Right(waleedk.medium.com)
2 points by Corgipower12 5y ago|0 comments
17.
by Corgipower12 5y ago
Is it always about things being faster? I would take less speed for better integrations and less development time.
18.
by Corgipower12 5y ago
This is a youtube video from this week's PyTorch Community Voices episode.
19.
Python to Distributed Python to Airflow task in ~5 lines of code(astronomer.io)
35 points by Corgipower12 5y ago|0 comments
20.
Ant Group's Resource Allocation System has scaled over 6000 cores(anyscale.com)
1 points by Corgipower12 5y ago|0 comments
21.
by Corgipower12 6y ago
Ray is a popular framework for distributed Python that can be paired with PyTorch to rapidly scale machine learning applications.
22.
Reinforcement Learning with the Unity Game Engine and RLlib(medium.com)
7 points by Corgipower12 6y ago|1 comments
23.
by Corgipower12 6y ago
Show how to integrate RLlib with the Unity Game Engine. Make the Soccer--Strikers vs Goalie training (or similar) complete faster by parallelizing a process.
24.
by Corgipower12 6y ago
In some organizations technology adoption decisions aren’t controlled by centralized teams or processes. As such different teams may end up independently choosing different cloud providers and their organizations end up multi-cloud. But whi
25.
by Corgipower12 6y ago
What I’d like to share with you in the next few sections is an example of code that was difficult to parallelize, but which Ray makes easy to parallelize. Two relatively straightforward refactors led to to a 60% improvement in performance o