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

Questions tagged [machine-learning]

Methods and principles of building "computer systems that try to automatically improve with experience."

0
votes
0answers
7 views

Statistical test to measure significant difference between 4 binary classifiers?

I have four versions of a neural network hardware. I am testing the hardwares to classify a single binary input into a binary output. The ANN hardwares are trained using a very simple relation, the ...
0
votes
0answers
36 views

My Neural Net can overfit but not generalize

I have created a Neural network that gets its training data from a complicated physics simulation. I run the simulation by randomizing 7 different inputs. Each input can be 1 of 4 discrete values. I ...
1
vote
0answers
15 views

Fitting a neural network with more parameters than observations

I'm training a neural network for regression using keras with about 13k training observations, each with 40 features. It's a Sequential model with Dense layers. I generate random architectures for ...
1
vote
1answer
26 views

Leave one out cross validation with classification - is that possible?

Doing "leave one out cross validation" with a regression task is easy. You can calculate the MSE (mean squared error) even on one single sample and average them. But what about a classification task? ...
1
vote
0answers
5 views

mlbench synthetic datasets for python

Is there any implementation of the synthetic datasets from mlbench in python? I am looking for these implementations in particular. I've found sklearn.datasets for synthetic examples, but they do not ...
1
vote
1answer
39 views

Integrating out parameter with improper prior

I got this problem while I was reading the book "Machine Learning: A Probabilistic Perspective" by Kevin Murphy. It is in section 7.6.1 of the book. Assume the likelihood is given by $$ \begin{...
0
votes
0answers
15 views

What is the difference between finding the threshold and training a binary classifier?

I have 1-D data: the output of a Cosine similarity (COS) distance between two features. The data is bounded by [-1,1]. When I just try to find the optimum threshold for class labels - I get an ...
1
vote
1answer
21 views

New factor levels in testing data set not present in training data in h20.randomforest

In randomforest classification using h20 package, there are factor levels which are present in testing data but not in training data.There is a warning message in predicting the values of testing data,...
0
votes
0answers
13 views

How to train a own question and answer dataset [on hold]

I have question and answer data set and i want to train a model so that similar questions will have same answer, How to train the data and which network should i use.. As classification may lead to ...
0
votes
0answers
7 views

DCGAN generator accuracy doesnt improve for high-res images

I trained a DCGAN on MNIST and CelebA dataset with 28x28 image size. Both the models were able to train successfully. I used many tips from https://github.com/soumith/ganhacks to make both the G and D'...
1
vote
0answers
12 views

Find roots of regression function

Physical problem: Signal data comes from several sensors (e.g. 4) and there is empirical knowledge, that data of one of them (e.g. "productivity") depends on other data. Signals of other sensors could ...
0
votes
1answer
30 views

How to deal with overestimation of small values and underestimation of high values in XGBoost?

I'm running XGBoost to predict prices on a cars dataset, I was wondering what alternatives are there for this kind of problem ...
0
votes
1answer
18 views

Why there is decrease in AUC values with increase in number of iteration in Held out validation?

I have a dataset with 600 rows and 4000 columns for which I am trying to do held-out cross-validation with 10 and 100 iterations. At first, the dataset is split into 80%:20% training and “held-out” ...
0
votes
0answers
7 views

PAC-Bayes bound for multiclass classification

I am starting to learn PAC learning, and have an interest in PAC-Bayes bound. However, most of the materials I found assumed binary classification only, while I am looking for the extension of PAC-...
0
votes
0answers
33 views

Generating vector image from a hand drawn picture. Machine Learning

I am new to machine learning! I need a way to generate vector image out of hand drawn sketch. I dont need to trace bitmap like it is usally done because it gives you exactly what you drawn. I need to ...
0
votes
0answers
24 views

What is the relationship between Online Learning and Statistical Learning?

Online Learning also known as Online Convex Opimization has famous algorithms like Follow-the-Leader and Online Gradient Descent (See OCO Book) Now stoastic programming has algorithms like Sample ...
0
votes
0answers
6 views

How to approach node/graph classification in an event?

I'm facing a new project and thought about maybe going in the direction of Graph-Neural-Networks. My data comes in the form of events (unrelated to each other), the data in each events contains a 2D/...
0
votes
1answer
20 views

What problem is it when I want to map these documents to these 3 different labels?

I am completely new to machine learning, and that means I am new to the ML-related jargon too. I have a problem at hand where there about a 1000 documents (on an average 500 words each) which need to ...
0
votes
0answers
10 views

Network Anomaly Detection based on conbinaison of two machine learning algorithms [on hold]

Bonjour a tout! J'ai un problème. Je suis un nouveau dans la détection des anomalies dans un réseau. Et c'est vrai que je souhaite utiliser le Kmeans et l'arbre de décision CART puisque j'ai lu ici ...
-2
votes
0answers
23 views

Reconstruct full Observations of a Time serie

I'm working on a project concerning a prediction of full observations applied to wind data (a time serie) which is called "Statistical Downscaling". From a time serie X which covers 20 years for ...
1
vote
0answers
16 views

Identifying the best fitting distribution type to test

Hi I tried to analyze a dataset, of USDT concurrency and following is the frequency distribution I received for log return I need this dataset to fit for a particular distribution. Can we suggest ...
0
votes
1answer
20 views

Driver Ranking based on Aggressive Cluster

I am clustering driver behavior data & found the most aggressive cluster where the means of all critical variables are high when compared to other clusters. Now can I take the agressive cluster &...
0
votes
0answers
10 views

Linear Discriminant Analyses (LDA)

in Linear Discriminant Analyses objective function, what is the effect of Trace? in other word what is the difference between (A'SWA)/(A'SWA) and Tr(A'SWA)/Tr(A'SWA) ??
2
votes
1answer
18 views

Can I apply Jensen's inequality in this case?

Suppose there is a function of two variables $f(X,\theta)$ which is convex with respect to $\theta$. Can I use the the Jensen's inequality in this case $E_{X}[f(x,\theta)]\geq f(E(x),\theta)$. The ...
2
votes
1answer
35 views

Save Machine Learning Model progress for later [on hold]

another dumb question, but how do you save the progress an ML model has made and start from that point later? Its kind of a vague question, but this is an example of what I am talking about: Say, ...
1
vote
2answers
26 views

Removing duplicates before train test split

Let's say you have a dataset generated from real world sampling which has lots of duplicates (the dependent and independent variables are identical) and you want to train a classifier to predict the ...
1
vote
1answer
43 views

Machine Learning techniques for uneven series of events

I don't know the correct terminology and so long everything I typed into google lead me to some version of time series modeling where all time series had the same number of points in a given time ...
0
votes
0answers
11 views

How do I update a Bernoulli prior parameter estimate after measuring additional covariates

I have a system that generates a probabilistic risk score, p0, for disease D0 from the results of an assay. The assay also generates several numeric features, f1, ..., fn, stored and available once ...
0
votes
1answer
9 views

Are there any studies of generalized error performance that don't assume data quality is constant with sample size?

As far as I know, much of the statistical and machine learning literature where modeling algorithms are compared for their generalization error performance as a function of sample size (think of ...
0
votes
0answers
11 views

Estimating true positive rate with a confidence interval given a classifier + unknown dataset truths

Let's say I've built a binary classifier - one that for instance, can classify whether a particular transaction is fraudulent or not. This classifier outputs a 1 or a 0, with a given degree of ...
0
votes
0answers
11 views

Why is XGBoost prediction proba so concentrated within specific range? (unbalanced class)

I am pretty to new to Machine Learning. I am training on some past Kaggle competitions including the Santander Customer Satisfaction Challenge (https://www.kaggle.com/c/santander-customer-satisfaction)...
1
vote
1answer
24 views

What happends if we standarise with a slightly shifted mean?

For deep neural networks or any ML algorithms , lets say the mean of my dataset is 5 and instead I normalize with 6 or 4, something slightly off. What affect does this have on learning?
2
votes
1answer
30 views

$L_2$ norm of product of two vectors

Let's assume we have two matrices $A^{d\times 1}$ and $B^{1 \times e}$, and we define their product as $C^{d\times e}$. Assuming $A,B$ are real valued with all entries in $[-1,1]$. I can intuitively ...
2
votes
1answer
23 views

K - means, expected shape of the curve [on hold]

I want to understand what happens as we increase the number of clusters using k- means, what is the expected shape of the curve showing the average distance between points and their assigned clusters? ...
1
vote
1answer
28 views

The proof for policy iteration algorithm's optimality

I am trying to understand why the policy iteration algorithm in Reinforcement Learning always improves the value function until it converges. Let's assume we have the policy $\pi_0(s)$ and our value ...
2
votes
0answers
26 views

How to fill NaN values that exist because there are no measures of certain features?

I'm currently doing a ML project (the goal is simply to clean the data set and apply some of the models we learned , like Random Forests, Ensemble learning, etc, and test the results) for a class and ...
2
votes
1answer
27 views

User-Specific Activity Level Along Weeks

I am assigning Low | Medium | High activity levels to users once a week. At the end of an entire period,a user has been assign n weekly activity levels among {Low, Medium, High}. Let k be the total ...
-1
votes
1answer
47 views

How to Identify dependent variable from my dataset?

I have a dataset which consists of 50 variables. I want to programmatically(using python) identify what is my dependent variable is. Are there any tests to separate the dependent variable from ...
4
votes
0answers
34 views

“Hierarchical” Random forests?

Background I am using Random Forest to classify ~900 objects based on a large number (> 80) predictors. I split these 70:30 for training and testing. The overall model does fairly well, giving an ...
2
votes
0answers
5 views

optimal scheduling allocation between online and walk-in queues

I have a scheduling queue that can be split between walk-ins and online booking appointments. In order to serve these queues I have a limited number of resources available. What is the optimal way to ...
0
votes
0answers
4 views

How to verify structural errors in a data set with huge number of variables

I have a data-set with 80 predictors and 1 outcome. As part of data cleaning i want to see the structure of each data, how they are distributed and whether there are any culprits(like NULL instead of ...
1
vote
0answers
28 views

How to predict routes using clustering data

I've been working on a ship route prediction algorithm such that given the past and current trajectory of a ship I am able to estimate the future one. The trajectories are represented as a sequence of ...
1
vote
1answer
47 views

how does node splitting occur in a decision-tree with non-categorical data?

According to a website (:http://dataaspirant.com/2017/01/30/how-decision-tree-algorithm-works/) , these values are chosen randomly: I don't think this is the case with internal algorithm of sci-kit ...
1
vote
2answers
53 views

How to extrapolate training metrics to test dataset?

i have split my training data 30/70 and trained models, my models are performing really well on the training set but i have a large unlabelled dataset where i want to do inference, how could i measure ...
1
vote
0answers
15 views

RMSProp Squared Gradients

In the RMSProp algorithm (And similar algorithms) that are used in Machine Learning in the subject of Adaptive Learning Rates, the squares of the gradients are used in the algorithm step. Is there ...
0
votes
0answers
5 views

Why goal of PLA can ignore the norm of normal vector

Define hyperplane $w*x+b=0$, the goal of PLA(Perceptron Learning Algorithm) is minimizing the distance of misclassified points to the decision boundary, i.e. $$-\frac{1}{||w||}\sum_{i\in M} y_i(w*x_i ...
1
vote
0answers
8 views

Learning Rate impact on model building time

I wanted to know that does learning rate impact the model building time in case of Gradient Boosted Trees. I do understand that increasing the number of trees have an impact( more the trees, more the ...
3
votes
0answers
29 views

Interpretation of box-counting method from R

I tried to calculate the fractal dimension of a dataset using the box-counting method with R programming. I used two packages: The first one is fractaldim, ...
3
votes
1answer
62 views

why has author divided by 1.5 in hands on machine learning with scikit learn

I am reading Hands-On Machine Learning with Scikit-Learn and TensorFlow (76/718), and the author is talking about dividing the dataset into a test set which i follow, but then goes on to talk about ...
1
vote
0answers
19 views

can you help me figure out this smart f1 function? [on hold]

Hi all this is a function being used to evaluate ML results y_true is the ground truth and y_pred are the predicted values from the machine learning model. This smart F1 function calculates the right ...