random-forest

I am new to this stuff and trying to attack Random Forest with Black Box FGSM (from clever hans) But I'm not sure how to implement it. They've a blackbox example for Mnist data but I dont understand where should I put my random forest and where should I attack. Any help would be appreciated. In the current tutorial, the black-box model is a neural network implemented with TensorFlow and its predictions (the labels) are used to train a substitute model (a copy of the black-box model). The substitute model is then used to craft adversarial examples that transfer to the black-box model. In your

I am currently using StringIndexer to convert lot of columns into unique integers for classification in RandomForestModel. I am also using a pipeline for the ML process. Some queries are How does the RandomForestModel know which columns are categorical. StringIndexer converts non--numerical to numerical but does it add some meta-data of somesort to indicate that it is a categorical column? In mllib.tree.RF there was parameter call categoricalInfo which indicated columns which are categorical. How does ml.tree.RF know which are since that is not present. Also, StringIndexer maps categories to

What am I doing? I am solving a classification problem using Random Forests. I have a set of strings of a fixed length (10 characters long) that represent DNA sequences. DNA alphabet consists of 4 letters, namely A , C , G , T . Here's a sample of my raw data: ATGCTACTGA ACGTACTGAT AGCTATTGTA CGTGACTAGT TGACTATGAT Each DNA sequence comes with experimental data describing a real biological response; the molecule was seen to elicit biological response (1), or not (0). Problem: The training set consists of both, categorical (nominal) and numerical features. It is of the following structure:

I am running a training of my model and I am getting the StackOverflow error whenever I increase the maxDepth over 12. Everything works correctly for 5,10,11. I am using spark 2.0.2 (and i cannot upgrade it for next couple of weeks). I have > 3M data, 200 features, 2500 trees and I would like to improve the accuracy by increasing the max depth. Is there a way to overcome this problem? Exception in thread "main" org.apache.spark.SparkException: Job aborted due to stage failure: Task 92 in stage 789.0 failed 4 times, most recent failure: Lost task 92.3 in stage 789.0 (TID 66903, 10.0.0.11): java

I am working on a Classification problem with 2 labels : 0 and 1. My training dataset is a very imbalanced dataset (and so will be the test set considering my problem). The proportion of the imbalanced dataset is 1000:4 , with label '0' appearing 250 times more than label '1'. However, I have a lot of training samples : around 23 millions. So I should get around 100 000 samples for the label '1'. Considering the big number of training samples I have, I didn't consider SVM. I also read about SMOTE for Random Forests. However, I was wondering whether NN could be efficient to handle this kind of

I fitted a random forest for my multinomial target with the randomForest package in R. Looking for the variable importance I found out permutation accuracy importance which is what I was looking for my analysis. I fitted a random forest with the h2o package too, but the only measures it shows me are relative_importance, scaled_importance, percentage . My question is: can I extract a measure that shows me the level of the target which better classify the variable i want to take in exam? Permutation accuracy importance is the best measure I can use in this case? For example: I have a 3 levels

This question already has an answer here: Scikit learn - fit_transform on the test set 1 answer I am currently trying to build a text classification model (document classification) with roughly 80 classes. When I build and train the model using random forest (after vectorizing the text into a TF-IDF matrix), the model works well. However, when I introduce new data, the same words that I used to build my RF aren't necessarily identical to the training set. This is a problem because I have a different number of features in my training set than I do in my test set (so the dimensions for the