Stack Exchange Network

Stack Exchange network consists of 174 Q&A communities including Stack Overflow, the largest, most trusted online community for developers to learn, share their knowledge, and build their careers.

Visit Stack Exchange

H2O is an in-memory platform for distributed, scalable machine learning.

0
votes
1answer
15 views

Issues with XGBoost on H2O environment

I have a dataset from which I built lags at different levels to use as features in the XGBoost model. When I ran XGBoost models on H2O, the model is picking up the features which contain higher values ...
1
vote
1answer
26 views

Is setting a self learning system possible via incremental (online) learning?

Self learning and incremental learning are all new to me. I am trying to develop a system for one of my case. Simply I have a data set (with about 90K observations and 400 features) for a binary ...
1
vote
2answers
38 views

How NULLs in numerical variables are treated in tree-based models?

I understand that in tree-based models (CART, Gradient boosted trees, etc.), NULLs (i.e., NaN) in categorical variables can be treated as a separated category, while making node splits. However, how ...
1
vote
2answers
37 views

h2o : Any components not running or stored on local machine? [closed]

The descriptive language used when referencing how h2o works is a little confusing to me at times (e.g., client, cluster, "internet"). I try to remain vigilant in not uploading or exposing my datasets ...
1
vote
1answer
45 views

Help about cosine simulating in h2o [closed]

So I'm beginning in deep learning and especially in h2o. I tried to simulate cosine function in R, not to compute it like for example by using h2o.cos(), But after many and many more combinations of ...
1
vote
0answers
28 views

balance_classes in H2O, but for regression? [closed]

I am training with deep learning for regression in H2O for R. My dataset is unbalanced (ie. not evenly distributed). There has been discussion on whether unbalanced datasets are an issue or not, with ...
4
votes
1answer
35 views

Special values in continuous numerical variables/features in Random Forest

I have a binary response variable I am seeking to predict using Random Forest. I have a sizable dataset of 150k rows, I have about 200 independent variables or features to use to model the outcome. ...
0
votes
1answer
34 views

Extracting H2o Cross Validation Results

I am using H2o library in R and have a slight confusion that you learned people might be able to help with. I am not sure how to interpret the output from ...
1
vote
1answer
31 views

H2O PCA number of components

I wonder why number of components in H2o PCA algorithm is limited to 9. It is not sure sometimes to be enough. k: Specify the rank of matrix approximation. This can be a value from 1 to 9 and ...
0
votes
0answers
42 views

How to improve GBM performance

I'm trying to model insurance losses, with a Tweedie distribution. I have a data set of about 40 million records, and over 100 independent variables. My response variable is "loss", I take the log of ...
2
votes
1answer
167 views

H2o interpretability - LIME

I have trained a model to predict heart attacks using random forest algorithm using H2O. I have good performance in cross validation. Now, I want to give more interpretation to the predictions in a ...
2
votes
1answer
61 views

How is deviance defined in H2O Deep Learning if the loss function is not quadratic?

One can set the parameter “stopping_metric” to “deviance” for the deeplearning algorithm (http://docs.h2o.ai/h2o/latest-stable/h2o-py/docs/modeling.html#h2o.estimators.deeplearning....
0
votes
0answers
25 views

How is the parameter „huber_alpha“ defined in H2O Deep Learning?

In the documentation of the H2O software (http://docs.h2o.ai/h2o/latest-stable/h2o-docs/data-science/deep-learning.html) it says for the parameter “huber_alpha: Specify the desired quantile for Huber/...
1
vote
1answer
306 views

H2O AUTOML: How to save, reuse and build on top of existing automl models? [closed]

I have two questions on h2o.automl and I couldn’t find any documentation on these topics. I can save/reuse the leader (automl) model in R using h2o.saveModel and h2o.loadModel. But how do I save/...
0
votes
1answer
91 views

feature engineering for auto encoder anomaly detection

I am working on an unsupervised problem: to take a set of transactional data, and identify anomalous transactions. I am using h2o's auto encoder to train a model which then scores transactions based ...
0
votes
0answers
30 views

“beta_given” column in the h2o.glm beta_constraints

What does the "beta_given" column do in the h2o.glm beta_constraints parameter? h2o is an open source library for machine learning algorithms. There are several online examples on how to install the ...
1
vote
1answer
30 views

Difference in parameters between grid summary and individual models

I was using h2o for getting a feel of what learning rate does a good job on my data set. I chose to use cross validation for selecting a learning rate. ...
1
vote
0answers
688 views

Implementing “Balanced Random Forest” in h2o

Currently having problems with (very) imbalanced classes in a dataset that I am trying to run binary classification on. Looking at this UC Berkeley article (http://statistics.berkeley.edu/sites/...
1
vote
1answer
93 views

Reduce over-forecasting in H2O regression with tree-based algorithms

I'm trying to solve a time series forecasting problem. While accuracy is fine, I'm now trying to punish / reduce under-forecasting events where the forecast is too low. Forecasting approach: Build ...
1
vote
1answer
402 views

Training threshold vs validation threshold for better prediction results?

Between the two, should I use a model's training or validation threshold to get best results (from a distributed random forest binary classifier built using h2o.ai) ...
5
votes
1answer
161 views

Deviances in H2O

does anyone know how exactly the Deviances (Poisson, Gamma, Tweedie) are computed in H2O? I cannot find the functions. For interpretation purposes I would like to know the calculations. Thank you!
4
votes
1answer
409 views

Regularized GLM with aggregated data

I am fitting a poisson GLM to model claim rates. Since I have 1.5m+ records, I have aggregated my data (to improve efficiency). My understanding is that using aggregated data with a poisson GLM will ...
0
votes
1answer
112 views

Which algorithm is used for the H2O AutoML metalearner [closed]

We know that h2o.automl() uses GBMs, Random Forest, DNNs, GLMs, and Extremely-Randomized Forest as base-learners, but what algorithm is used as the metalearner to ...
1
vote
1answer
454 views

Can we force monotonicity for a subset of features in h2o gbms?

I was wondering if it's possible to force variables to be monotonic when building gbms in h2o? Ideally this would be enforced globally, although even the standard local constraint would be better than ...
2
votes
1answer
161 views

Is polynomial regression possible in H2O? [closed]

Is there a way to carry out polynomial regression $x + x^2$ in H2O (Python)? What I have found about this is "interactions" option in GLM. However, I am not sure if this option yields polynomial ...
0
votes
2answers
412 views

can I use h2o or mxnet to do time series prediction?

I have a time series like 4, 6, 10, 12, 2, 23, 4,... It is not a stationary time series so I do not want to use arima algorithm. I perfer neural networks. There is some algorithm in R package such as ...
2
votes
1answer
313 views

Random Forest cross-validation r2 is high but predictions on simulated data are bad

I have a dataset with 46 million observations and 25 predictors. I am training my model in the python h2o package like so: ...
1
vote
1answer
129 views

How are the epochs and iterations defined in a h2o deep learning perspective?

As per the usual definition of epoch and iteration, it takes multiple iterations to complete 1 epoch (for example see https://stackoverflow.com/questions/4752626/epoch-vs-iteration-when-training-...
0
votes
1answer
248 views

Use of previously set cross validation set for ensemble models in h2o

I have time series dataset and as a requirement I prepare a 10 fold cross validation dataset. Briefly, I split data into 10 and for each training set I put a gap between last, first day of validation ...
2
votes
1answer
1k views

Random forest variable importance in h2o (classification problem)

I cannot find out how the variable importance for classification problems is calculated in h2o. There is a Stackoverflow question asking the same, but the accepted ...
1
vote
0answers
36 views

When to use different kinds of performance metrics [closed]

In h2o.ai you can get the threshold for (binary classification) models that maximize different metrics (“min_per_class_accuracy”, "mean_per_class_accuracy", “...
0
votes
0answers
23 views

Training errors bigger than validation errors

I'm using H2O distributed random forest for a binary classification problem of a dataset with mix categorical and real features. However, my validation set errors are smaller than the training set ...
2
votes
1answer
300 views

How to map values reconstructed with a deep learning autoencoder back to categories

My colleague and I are working on a dataset full of categorical variables using an deep learning model with autoencoder of H2O package in ...
1
vote
0answers
152 views

Splitting criteria based on MSE in H2O DRF (Random Forest) and GBM

Why do H2O Random Forest and GBM implementation use MSE (mean squared error) for node splitting metric? As I understand from the H2O docs, no matter the response column is factor or numeric, the tree ...
1
vote
1answer
392 views

L1/L2 Generalization with value setting 0 in H2O

I'm am using the h2o.deeplearning() function in R, and in the parameter setting, there is l1 and l2. The documentation for L1/L2 regularization says: L1 ...
2
votes
1answer
109 views

Fractional epoch in deep learning using H2O

I understand the concept of epoch as epoch = one forward pass and one backward pass of all the training examples, in the neural network terminology. So i assume it must be an integer, however why in ...
3
votes
1answer
165 views

use of sample_rate = 1 in randomForest - to fit a single tree [closed]

I would like to fit a single tree. In the h2o R package, I can use h2o.randomForest() with the following options: ...
3
votes
1answer
950 views

With H2O AutoML is it okay to use my test set as the leaderboard?

Normally in machine learning we will split our data into train, valid and test. The valid data is used to tune the parameters, and the test data is then used to check the performance of our best tuned ...
0
votes
2answers
287 views

machine learning model true negative rate is too low while true positive rate is too high

I am using the tm package and h2o package to do text mining using neural network. Here I have a data frame of 100 most frequent words in the text. These variables only have values of non-negative ...
1
vote
1answer
195 views

Sudden increase in log-loss on training set with H2O GBM

I am performing a hyperparameter grid search for a GBM classifier in H2O, running version 3.10.4.8 (on top of python 3.5.3). This is a multiclass problem (~40 classes). As a first test, I tried a ...
0
votes
1answer
393 views

Why is my predicted vs observed plot worse for training than validation. Running an overfitted GBM on a binomial outcome

I have a binomial outcome that I am trying to predict using a gbm in h2o. I have set quite a low min_rows value for each node and it appears to be overfitting. See plots below. When I group the ...
2
votes
2answers
3k views

H2O: Can I use the h2o for time series predictions?

I understand that there is not a specific model for time series modeling in H2O. Is there a workaround in order to use Deep Learning or/and GBM? Is some kind of data transformation necessary? are ...
1
vote
1answer
127 views

Neural Network and H2O.ai: inputs that have multiple right answers (questions about a previously answered question)

I have two questions related to this previously answered question. Can someone explain how the NN's backproagation algorithm would not be able to function properly when inputs have multiple right ...
1
vote
1answer
68 views

Is it necessary to logarithm continuous attributes, when tree construction is histogram powered?

LightGBM wiki says: LightGBM uses the histogram based algorithms, which bucketing continuous feature(attribute) values into discrete bins, to speed up training procedure and reduce memory usage. ...
2
votes
1answer
392 views

How does h2o handle time-series cross validation?

I've read about How does h2o.r cross validation work?. However, for a time series dataset, does H2o support the type of CV described here Using k-fold cross-validation for time-series model selection? ...
0
votes
1answer
826 views

How the Internal H2O auc measures are calculated? Why they are so close to 1 or 1?

I am randomly holding out 10% of data out of the whole dataset as test.data and train the GBM model on a remaining 90% of rows train.data (with x and y provided, no nfolds or validation data set ...
0
votes
1answer
1k views

How does h2o.r cross validation work?

For GBM and randomForests: I understand, that when I set nfolds to 10, the training frame is divided into 2 sets, first having 90% of rows, and second having 10% of rows randomly for first cross ...
0
votes
1answer
236 views

Tuning Binary PCA using Generalized Low Rank Model

I am using h2o Generalized Low Rank Models for tuning a binary PCA (Logistic loss + Quadratic regularization) to perform a recommender system task. I am using the H2O infrastructure using the GLRM ...
0
votes
0answers
42 views

How to choose relevant category for metrics in h2o in R?

I have problem with h2o metrics during cross-validation. I have binary classification task with huge overrepresentation of one class. When I preform learning h2o generates metrics for every fold. ...