And introduce an abs(TC) too please…
what the overall impact will be on the metamodel
Looking forward to higher PF
So beautiful! The core idea sounds weirdly obvious in hindsight, but undoubtably tough to come up with and implement.
Since stake size plays a significant role before regularization, will models with large stakes (1+% of metamodel) have more volatile metrics on TC or even be at a disadvantage? Or does the regularization reduce almost all effect of stake size?
Would you recommend large stakers to spread out the stake over more (diverse) models when optimizing for TC?
I hope we can have these metrics’ importance w.r.t. TC in every resolved round to see the dynamics. I am curious about the new TC stacking options, how will it affect mmc importance as the meta-model may become more different from the example model. I am still not sure if i should optimize for mmc directly by the example prediction.
I am very excited for these changes. Even if there is a slight dip in NMR value, this is a big step forward for the hedge fund which we all want and need to stick around for a long time :). Also from a profitability stand point, I appreciate that the barrier to entry is rising and that it is strategy related (feature engineering and modeling to optimize TC) not necessarily hardware dependent (looking at you supermassive dataset).
I am wondering though if the team has done any testing and experimentation with the validation set and optimizing TC? Was the above analysis done inclusive of validation eras?
I have mixed feelings about the current validation set for Corr and MMC and so wondering if there are any related changes or improvements down the pipe in this regard?
I think I answered this here: Question on TC: Is it True Contribution or something else? - #3 by mdo
no doubt in the end this will happen
@mdo Assume someone stake 100 NMR , what’s the differences between stake on (0 x CORR+2xTC) and (1xCORR+2xTC) ?
Is this new metric TC robust towards p / (1-p) type of vulnerability?
this makes sense. Outliers like Round 304 Alfaprism_41 corr and mmc in 99 pct and TC in 25pct needs some unpacking.
It is symmetrical (1-p will get exactly opposite TC), and there is no bonus anymore, so shouldn’t be an issue.
Is there any preprocessing for user’s predictions before SWMModel in the production system? In MMC, user’s predictions are converted to uniform distribution but I am wondering the behavior for TC.
Especially whether only rank matters or magnitude matters for TC calculation, and this information is useful for us to understand the TC behavior.
For those who are commenting that they will likely withdraw from the competition due to the change as their currently optimized model (CORR/MMC) is not suitable for TC, I think that is their intention. Because those models are earning rewards without helping TC/hedge fund performance.
I am fortunately not in that group somehow. I recently checked my model and it would have made 1% more per week if I could stake on 2xTC instead of 2xMMC. It does however increase volatility as others have mentioned, but perhaps for a different reason. I currently have a 2.1 Sharpe with Corr+2xMMC but it will drop to 1.3 with Corr+2xTC. This is mainly due to the lack of correlation between CORR and MMC in my submissions, but a 40% correlation between CORR and TC. Ideally, having some validation diagnostics to historically backtest TC will be helpful if that can be provided in the diagnostic tool.
train_example_preds.parquet file, at some point, be provided? If we are interested in modeling an optimization for
Exposurer Dissimilarity (mentioned in the original post from @mdo )? For those willing/wanting to build a new tran/validation set via mixing eras from each, it becomes impossible to draw from an existing
Would a relatively decent equivalent be to use the example model provided to produce predictions for the training set? Or are the
example_preds in the validation file specially chosen by the Numerai team for a particular reason?
lol well I will stay skeptical initially and see what will happen with TC rankings and daily/weekly scores for example, I can imagine we will see some huge swings when people start experimenting. At the moment I have a model ranked #11 for TC, yet I have absolutely no clue why this one would be up there so high in the ranking
Time to get staking! but before you do, tell me about your model…
lol nothing fancy here, its an ensemble of different regression algo’s with the small feature selection, and also only 1/4 of the v2 dataset (1 out of 4 era’s to avoid overfitting).
you know, i tried something like that with ElasticNet, Ridge, Lasso and Lars. Nil Nada Niet.
I hope you’re gonna sell those predictions on numerbay, fame and fortune beckons… if you’re in the market for an agent…
Well… be my guest I would say : https://numerbay.ai/product/numerai-predictions/bigcreeper_4
Maybe this has been already answered somewhere else and I missed that, but How is that the documentation and diagnostic tool still mention
mmc instead of
tc? Is there a way to know the
tc for models not submitted ?