Questions tagged [machine-learning]

Machine learning algorithms build a model of the training data. The term "machine learning" is vaguely defined; it includes what is also called statistical learning, reinforcement learning, unsupervised learning, etc. ALWAYS ADD A MORE SPECIFIC TAG.

Filter by
Sorted by
Tagged with
1
vote
0answers
23 views

Why do the XGBoost predicted probabilities of my test and validation sets look well calibrated but not for my training set?

I am using an XGBoost classifier to predict propensity to buy. After drawing a calibration curve to check how well the classification probabilities (predict_proba) produced are vs actual experience, I ...
0
votes
2answers
163 views

Multi-step ahead forecasting with LSTM neural network

I would like to forecast the heat load of a district heating network given its past values, the temperature and the 3-day ahead forecast of the temperature with an LSTM RNN. The data is hourly and I ...
3
votes
3answers
2k views

Dealing with new Factor Levels in a Regression in R

I originally posted this in stackoverflow (as given here) but was told to try here since it might be more relevant here. I am very new to statistics and R in general so my question might be a bit dumb,...
1
vote
1answer
29 views

Text classification with small dataset for a specialized domain

I have a multiclass text classification problem where I have very few documents for each class. The classes are imbalanced but I want to be able to predict the class when I have at least 200 - 300 ...
14
votes
2answers
3k views

Run-time analysis of common machine learning algorithms

Does anyone have reference to a summary of run-time analyses for common machine learning algorithms (different flavors of NN, SVMs, etc)?
3
votes
2answers
106 views

Do I need 3 RGB channels for a spectrogram CNN?

I am computing a linear spectrogram of an audio signal. https://en.wikipedia.org/wiki/Spectrogram The spectrogram is a 2-D matrix with time on the x-axis and frequency on the y-axis. The traditional ...
0
votes
1answer
20 views

Interpreting trees of XGBoostRegressor Model

I fitted a dmlc XGBoostRegressor model on a dataset with the variables ['CPI', 'Fuel_Price', 'Temperature', 'Unemployment'] and ...
0
votes
1answer
31 views

How to deal with biased dataset for both training and testing data?

I am currently working on a classification problem with a highly biased dataset. The dataset is biased for both training and testing data. And I am having trouble dealing with the dataset or modifying ...
0
votes
1answer
471 views

scale factor for MAD for non-normal distribution

I understand that the scale factor for normally distributed data is 1.4826 to convert it to a pseudo standard deviation like quantity which could be used with the median for determining confidence ...
0
votes
2answers
42 views

Is it possible to know if a machine learning model is overfitted from negative samples?

I have a trained model and with this I can classify faces, if I test the classifier by entering the same negative samples (not faces) with which I train, is it possible to know if my model is ...
0
votes
0answers
32 views

Normalizing results of a probabilistic classifier

I have a probabilistic classifier that produces a distribution over my 3 classes - C1, C2, C3. I want to compare some new points I'm classifying to each other, to see which one is the best fit for a ...
3
votes
1answer
333 views

Combining classification and anomaly detection

I want to build a system, that can classify known classes in a supervised way and at the same time tells if there is a new anomaly class it has not seen before. The user can then label that unknown ...
0
votes
0answers
14 views

Object distortion after ROI Align in Mask R-CNN

In Mask R-CNN, if there are 2 proposed ROIs which cover 2 objects that looks like below: #1 A square object #2 A rectangular object So my question is: After ROI Align, is the #2 feature map ...
1
vote
1answer
144 views

Feature selection in scikit-learn and multilinear regression

I have a process with ~10 features and ~100 responses, and would like to search for models for how those features interact to create various responses. ~100 experiments were run, exploring ...
1
vote
1answer
1k views

Stacking models which trained by different features in a data set for a classification problem

Normally, the first layer models in stacking and bledning method are trained by all features in a data set. However, what will be the performance of a big model based on first layer models trained by ...
2
votes
0answers
22 views

Understanding the trade-off between bias and variance in machine learning prediction using the math formula

About machine learning prediction, I would like to understand the trade-off between the bias and variance but using the mathematical formula. We have some train data X and y target variable. To be ...
3
votes
1answer
5k views

How to Compare the Data Distribution of 2 datasets?

I'm having trouble to understand how to compare 2 sets of data by their distribution . For Example, how can I understand that column X100 has the same distribution as column Y1? Also, is there a ...
1
vote
2answers
40 views

Intuitive Understanding of Expected Improvement for Gaussian Process

So I am learning Bayesian Optimization and came across expected improvement. My question is are we searching for the point in the Gaussian Process model whose expected value (determined by mean and ...
1
vote
1answer
278 views

How does word2vec work for word similarity?

I am trying to apply word2vec/doc2vec to find similar sentences. First consider word2vec for word similarity. What I understand is, CBOW can be used to find most suitable word given a context, whereas ...
1
vote
1answer
143 views

What would be the final hypothesis like? if our unknown target is a distribution rather than a function?

The above picture is about building model, it seems a bit difficult to understand the meaning of "plus noise", and what would the final hypothesis look like? if the unknown target changes from ...
0
votes
1answer
40 views

Bayes formula alternate expression using alpha

I know that Bayes theorem is: Posterior = Likelihood * Prior / Evidence However, I am confused about the above notation in the picture. How do we get to the above three notation? How does ...
3
votes
1answer
58 views

How did these researchers determine the confidence interval of the AUROC using resampling but without retraining the model?

In this Nature article backed by Google, the investigators develop then externally validate a deep learning model for predicting lung cancer using CT scans. In their internal validation results, we ...
0
votes
0answers
17 views

Draw chance in games as a range?

I am working on my first project with a lot of data and outcome prediction. One goal is to calculate the expected outcome of a game between two parties. (Parties don't have to be of equal "strength" ...
7
votes
4answers
8k views

What is the difference between independent variable and a feature?

I ran into this question which asks the identification of various terms for a linear regression function (f). I am confused about the "independent variable" definition. What is the difference ...
0
votes
1answer
15 views

Machine learning on Percent as dependent variable

I am working on a problem where I am tasked to predict users into 'High users' and 'Low users'. Dependent variable provided is in percent of orders (%) which is calculated as (#orders placed/#sales ...
0
votes
1answer
452 views

Machine Learning on Percent/Continous Dependent Variable

I have a large dataset of 30,000 cases with 150 variables. I am looking for a few possible machine learning solutions/methods that I could try and use for cross validation. My dependent variable ...
1
vote
1answer
1k views

guide for text classification using weka

I have a set of 2000 small texts (each less than 500 words) that I manually categorized. All the texts are in the same main subject, and I want to separate them into distinct groups based on their ...
16
votes
1answer
2k views

Why isn't Akaike information criterion used more in machine learning?

I just ran into "Akaike information criterion", and I noticed this large amount of literature on model selection (also things like BIC seem to exist). Why don't contemporary machine learning methods ...
4
votes
2answers
70 views

Are loss functions necessarily additive in observations?

In all of the contexts I've seen loss functions in statistics/machine learning so far, loss functions are additive in observations. i.e.: loss $Q_D$ of dataset $D$ is an additive aggregation of losses ...
0
votes
0answers
12 views

How do distance-based models like Logisitic regression handle features with zero values? [on hold]

Distance based-models will always attribute a zero-importance of the zero values of samples (like categorical features with 0 and 1) in the optimization process. For example, for a logistic ...
11
votes
2answers
2k views

Use Pearson's correlation coefficient as optimization objective in machine learning

In machine learning (for regression problems), I often see mean-squared-error (MSE) or mean-absolute-error (MAE) being used as the error function to minimize (plus the regularization term). I am ...
2
votes
2answers
802 views

Nonlinear Dynamic Online Classification: Looking for an Algorithm

I have two predictors a,b that I want to use combine to classify data. a is stable, it will always produce the same prediction for the same input. b will change and probably improve in time (because ...
0
votes
0answers
38 views

I want to know the relationship between Discriminant functions and the kernel in SVM

The following articles are reprinte of #3338212 of math.stackexchange.com. It was recommended to ask this community at math.stackexchange.com. The following 【Quiz】 and 【Official Answer】are the ...
0
votes
0answers
5 views

What is gnmt, mask, resenet, transformer in MLperf result? [on hold]

https://github.com/mlperf/training_results_v0.6/tree/master/Google/benchmarks We can see these results, however, I can not map them to their repo: https://github.com/mlperf/training
3
votes
1answer
884 views

How to use weights with Elasticnet regression in python?

I am using Elasticnet from scikit-learn in python, I've also used Glmnet package in R for prototyping. I want to use weights in Elasticnet which apparently is not available as an option/argument in ...
0
votes
0answers
7 views

How can I detect multiple features from a video stream

I want to detect multiple things when a person steps into view of the camera. Their face Their badge How would I go about implementing a classifier to detect two objects and put them together into ...
0
votes
0answers
16 views

Rewrite Wilk's Lambda using Kernel function

I am new to the data learning field. I am trying to learn more about kernels and kernel function. I understand how to generate a kernel, however, recently I was trying to use kernel functions in ...
0
votes
1answer
52 views

KL Divergence of two standard normal arrays

I generated two 9000,1 np arrays with a = np.random.standard_normal(9000) b = np.random.standard_normal(9000) Then I check the KL Divergence with ...
1
vote
1answer
39 views

Should exploratory data analysis include validation set?

I know that EDA should be performed on the training set but not on the test set. But my question is: we usually split the training set into training and validation datasets. Should we perform EDA on ...
8
votes
2answers
1k views

Scaling the backward variable in HMM Baum-Welch

I am just trying to implement the scaled Baum-Welch algorithm and I have run into a problem where my backward variables, after scaling, are over the value of 1. Is this normal? After all, ...
0
votes
0answers
12 views

regression method to predict differences between A and B given P(A>B) as prior probabilities

First off I apologize if this has been asked before, I tried searching for came up empty handed. Please forgive me if this is duplicated. I have 2 streams of data A and B which are pricing data, each ...
0
votes
1answer
632 views

What is an autoregressive decoder?

I saw that this was part of a deep belief network I was looking at. I'm not sure what it means. Is it a layer that transforms few inputs into many outputs and has a connection to itself? What is an ...
1
vote
1answer
63 views

How is TD(1) of TD(lambda) equivalent to Monte Carlo?

In Sutton and Barto's book about RL they say that the TD($\lambda$) algorithm is equivalent to Monte Carlo when $\lambda = 1$. I don't see how that is the case. They define the lambda return as: $$...
0
votes
1answer
26 views

My approach to deal with the effects of randomness in a ML model

I'm not a statistician, so please excuse a possibly wrong use of terminology here. My dataset has about 400 - 800 samples and about 800 features. The samples are ordered by time, although it is not a ...
1
vote
1answer
20 views

Why we fit xᵢ vs errorᵢ in Gradient Boosting

The basic idea of Boosting is to reduce bias by reducing training error in multiple iterations. However, I'm unable to understand how does combining multiple models which are trained by fitting ...
0
votes
0answers
8 views

EPE for a categorical variable

I am going through "Elements of Statistical Learning", and my memory of statistics is a bit rusty at this point, so I have trouble understanding the following equation from the book: $$EPE = E_X \...
0
votes
1answer
25 views

Is increasing the class weight of minority class in Random Forest algorithm decreasing the performance?

I am trying to classify an imbalanced dataset (census dataset with approx. 3:1 imbalance) using Random Forest algorithm in python, and what I observed that increasing the class weight of the minority ...
0
votes
0answers
20 views

In semantic segmentation using Fully convolutional networks, why is Cross Entropy loss preferred over L1 or L2 losses?

I am training a fully convolutional network with Encoder-Decoder architecture for the task of Image Segmentation and currently am using the Binary Cross Entropy loss for foreground/background ...