human_scientist

6 ms·

1.
by human_scientist 7y ago
> To submit to arxiv you need to be approved by someone as a legitimate researcher, or beg for reviews from people you’ve never met without any anonymity. I did not have to do something like this.
2.
Automatic Machine Learning Tutorial [NeurIPS 2018](nips.cc)
15 points by human_scientist 7y ago|0 comments
3.
by human_scientist 7y ago
It depends on the number of evaluations. The more evaluations the stronger the model built by the TPE algorithm. With very few evaluations, we would expect TPE to match random sampling. This effect can for example be seen in the plots of th
4.
by human_scientist 7y ago
HyperOpt is a library for hyperparameter optimization, my comment was about algorithms. From the homepage of hyperopt: "Currently two algorithms are implemented in hyperopt: Random Search Tree of Parzen Estimators (TPE)" (TPE is a
5.
by human_scientist 7y ago
What is the expected lifetime?
6.
by human_scientist 8y ago
Nice project! You use gridsearch for hyperparamter optimization and state that at some point you would like to add a bayesian approach. One simple change that could boost the performance, would be to use random search inplace of gridsearch.
7.
by human_scientist 8y ago
The NFL only applies to settings where your task distribution is uniform random over all possible tasks. It is my intuition that this kind of task distribution is almost surely not something we would encounter.
8.
by human_scientist 8y ago
Your algorithm description sounds like a variant of greedy local search with a NN heuristic. While I personaly do not know of a paper that does this kind of local search (I have not looked), I do want to add that there is a rich literature
9.
by human_scientist 8y ago
> I don't see "Automatic design of novel algorithms" in this list. Can AutoML produce something as novel as a GAN, CapsNet, WaveNet, Transformer, Neural ODE, etc? Is that even considered to be one of its goals. In my opini
10.
by human_scientist 8y ago
The field of automatic machine learning (abbreviated as AutoML) concerns all endeavours to automate the process of machine learning. To provide a sense of what could constitute AutoML, let me post a list from the "Call for Papers"
11.
by human_scientist 8y ago
Why? The concept of AutoML does include the design of novel algorithms.
12.
by human_scientist 8y ago
Could you give an example? I have a hard time understanding what you could mean, as Algorithm Configuration & Selection is such a general framework. If you are solely talking about the current state of the art, I would agree that techni
13.
by human_scientist 8y ago
The parent did not specifically talk about NNs. As I understand it AutoML could apply to all statistical endeavours that involve estimation (classical or bayesian).
14.
by human_scientist 8y ago
Automatically building a scikit learn estimator might include many conditional hyperparameters and also a very large amount of them (<100) [1]. However, performing joint architecture and hyperparameter search can be framed to be on a muc
15.
by human_scientist 8y ago
What about approaches like auto-sklearn [1]? With these it is basicaly: >>> automl = autosklearn.classification.AutoSklearnClassifier() >>> automl.fit(X_train, y_train) >>> y_hat = automl.predict(X_test)
16.
by human_scientist 8y ago
https://eternagame.org/ This is RNA Design for humans as a game! Also: A trained human outperforms current AI approaches for RNA Design.
17.
Learning to Design RNA(openreview.net)
36 points by human_scientist 8y ago|2 comments