-
```
def logDensityOfCounts(self,aux):
term1 = scipy.special.gammaln(self.alpha) - scipy.special.gammaln(self.alpha + aux.numCustomers)
term2 = aux.numTables + math.log(self.alpha + (aux.numT…
-
Deep learning and NLP for higher text representations
http://colah.github.io/posts/2014-07-NLP-RNNs-Representations/
-
This is as much of a feature request as anything. In our science applications we often deal with datasets that have been masked for some reason (e.g. some observational limitation resulting in a selec…
-
Running this tutorial http://scikit-learn.org/stable/auto_examples/mixture/plot_gmm.html with verbose=True for DPGMM outputs large positive numbers for variational lower bound:
```
Bound after updati…
-
http://www.weibo.com/3774704830/BrpAC2mPE
-
我的问题是关于分类模型(classification model / framework)的,有时候可以对同一样本提取多种不同特征,一般来说,我们都假设特征之间是统计独立的,然后利用这些特征去做分类。
但我现在需要考虑的情况是,两个特征之间是非独立的,也就是说,其中一个特征是基于另一个特征所提供的信息而提取得到的,比如,在一幅图像中,先获取物体的形状(shape)或轮廓,再得到其中的外貌(app…
-
This is more a small brainstorming than a coding issue.
I tried to make a demo of the simplest machine learning situation that captures our problem of regression on noisy data. Assume `p` is a parame…
-
In python world, there is a very well known lib called [fuzzywuzzy](https://github.com/seatgeek/fuzzywuzzy). A nice description of its features is available [there](http://chairnerd.seatgeek.com/fuzzy…
-
私信
-
http://www.weibo.com/1974787502/BkfWmyZUV