Questions tagged [machine-learning]

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

Filter by
Sorted by
Tagged with
32
votes
6answers
6k views

How to set the number of neurons and layers in neural networks

I am a beginner to neural networks and have had trouble grasping two concepts: How does one decide the number of middle layers a given neural network have? 1 vs. 10 or whatever. How does one decide ...
34
votes
5answers
33k views

Does gradient descent always converge to an optimum?

I am wondering whether there is any scenario in which gradient descent does not converge to a minimum. I am aware that gradient descent is not always guaranteed to converge to a global optimum. I am ...
27
votes
6answers
10k views

Why do convolutional neural networks work?

I have often heard people saying that why convolutional neural networks are still poorly understood. Is it known why convolutional neural networks always end up learning increasingly sophisticated ...
150
votes
5answers
87k views

What is the “dying ReLU” problem in neural networks?

Referring to the Stanford course notes on Convolutional Neural Networks for Visual Recognition, a paragraph says: "Unfortunately, ReLU units can be fragile during training and can "die". For ...
68
votes
7answers
79k views

Open source Anomaly Detection in Python

Problem Background: I am working on a project that involves log files similar to those found in the IT monitoring space (to my best understanding of IT space). These log files are time-series data, ...
56
votes
5answers
30k views

Adding Features To Time Series Model LSTM

have been reading up a bit on LSTM's and their use for time series and its been interesting but difficult at the same time. One thing I have had difficulties with understanding is the approach to ...
50
votes
4answers
50k views

Why mini batch size is better than one single “batch” with all training data?

I often read that in case of Deep Learning models the usual practice is to apply mini batches (generally a small one, 32/64) over several training epochs. I cannot really fathom the reason behind this....
51
votes
5answers
12k views

Should I go for a 'balanced' dataset or a 'representative' dataset?

My 'machine learning' task is of separating benign Internet traffic from malicious traffic. In the real world scenario, most (say 90% or more) of Internet traffic is benign. Thus I felt that I should ...
132
votes
5answers
150k views

The cross-entropy error function in neural networks

In the MNIST For ML Beginners they define cross-entropy as $$H_{y'} (y) := - \sum_{i} y_{i}' \log (y_i)$$ $y_i$ is the predicted probability value for class $i$ and $y_i'$ is the true probability ...
50
votes
7answers
49k views

In supervised learning, why is it bad to have correlated features?

I read somewhere that if we have features that are too correlated, we have to remove one, as this may worsen the model. It is clear that correlated features means that they bring the same information, ...
44
votes
8answers
46k views

Why should the data be shuffled for machine learning tasks

In machine learning tasks it is common to shuffle data and normalize it. The purpose of normalization is clear (for having same range of feature values). But, after struggling a lot, I did not find ...
31
votes
4answers
13k views

Quick guide into training highly imbalanced data sets

I have a classification problem with approximately 1000 positive and 10000 negative samples in training set. So this data set is quite unbalanced. Plain random forest is just trying to mark all test ...
88
votes
8answers
109k views

When should I use Gini Impurity as opposed to Information Gain?

Can someone practically explain the rationale behind Gini impurity vs Information gain (based on Entropy)? Which metric is better to use in different scenarios while using decision trees?
25
votes
3answers
42k views

Data Science Project Ideas [closed]

I don't know if this is a right place to ask this question, but a community dedicated to Data Science should be the most appropriate place in my opinion. I have just started with Data Science and ...
24
votes
4answers
8k views

Role derivative of sigmoid function in neural networks

I try to understand role of derivative of sigmoid function in neural networks. First I plot sigmoid function, and derivative of all points from definition using python. What is the role of this ...
30
votes
4answers
12k views

Is it always better to use the whole dataset to train the final model?

A common technique after training, validating and testing the Machine Learning model of preference is to use the complete dataset, including the testing subset, to train a final model to deploy it on, ...
12
votes
1answer
12k views

How to Predict the future values of time horizon with Keras?

I just built this LSTM neural network with Keras ...
6
votes
4answers
4k views

How to give name to topics created using LDA?

I have categorized 800,000 documents into 500 categories using the Mahout topic modelling. Instead of representing the topic using the top 5/10 words for each topics, I want to infer a generic name ...
55
votes
10answers
51k views

Machine learning - features engineering from date/time data

What are the common/best practices to handle time data for machine learning application? For example, if in data set there is a column with timestamp of event, such as "2014-05-05", how you can ...
44
votes
2answers
39k views

How to interpret the output of XGBoost importance?

I ran a xgboost model. I don't exactly know how to interpret the output of xgb.importance. What is the meaning of Gain, Cover, and Frequency and how do we ...
17
votes
1answer
25k views

Why ReLU is better than the other activation functions

Here the answer refers to vanishing and exploding gradients that has been in sigmoid-like activation functions but, I guess, Relu...
24
votes
2answers
28k views

Should we apply normalization to test data as well?

I am doing a project on author identification problem. I had applied the tf-idf normalization to train data and then trained a svm on that data. Now when using the classifier should I normalize test ...
19
votes
1answer
17k views

back propagation in CNN

I have the following CNN: I start with an input image of size 5x5 Then I apply convolution using 2x2 kernel and stride = 1, that produces feature map of size 4x4. Then I apply 2x2 max-pooling with ...
25
votes
2answers
24k views

Why use both validation set and test set?

Consider a neural network: For a given set of data, we divide it into training, validation and test set. Suppose we do it in the classic 60:20:20 ratio, then we prevent overfitting by validating the ...
16
votes
3answers
6k views

Linear regression with non-symmetric cost function?

I want to predict some value $Y(x)$ and I am trying to get some prediction $\hat Y(x)$ that optimizes between being as low as possible, but still being larger than $Y(x)$. In other words: $$\text{cost}...
9
votes
2answers
6k views

Validation vs. test vs. training accuracy. Which one should I compare for claiming overfit?

I have read on the several answers here and on the Internet that cross-validation helps to indicate that if the model will generalize well or not and about overfitting. But I am confused that which ...
5
votes
2answers
6k views

What are useful evaluation metrics used in machine learning

I am using CNN in order to predict codes after analyzing text. As an example, I will write "I am crazy" .. the model will predict some code " X321". All this based on CNN. I want to evaluate my ...
8
votes
1answer
13k views

Why should softmax be used in CNN

In the last layer of CNNs and MLPs it is common to use softmax layer or units with sigmoid activation functions for multi-class ...
18
votes
2answers
13k views

Why do we need to discard one dummy variable?

I have learned that, for creating a regression model, we have to take care of categorical variables by converting them into dummy variables. As an example, if, in our data set, there is a variable ...
9
votes
1answer
2k views

Can The linearly non-separable data be learned using polynomial features with logistic regression?

I know that Polynomial Logistic Regression can easily learn a typical data like the following image: I was wondering whether the following two data also can be ...
142
votes
17answers
118k views

Best python library for neural networks

I'm using Neural Networks to solve different Machine learning problems. I'm using Python and pybrain but this library is almost discontinued. Are there other good alternatives in Python?
109
votes
16answers
101k views

How do you visualize neural network architectures?

When writing a paper / making a presentation about a topic which is about neural networks, one usually visualizes the networks architecture. What are good / simple ways to visualize common ...
123
votes
6answers
167k views

How to draw Deep learning network architecture diagrams?

I have built my model. Now I want to draw the network architecture diagram for my research paper. Example is shown below:
24
votes
6answers
9k views

Deep learning basics

I am looking for a paper detailing the very basics of deep learning. Ideally like the Andrew Ng course for deep learning. Do you know where I can find this ?
42
votes
3answers
62k views

How to set batch_size, steps_per epoch and validation steps

I am starting to learn CNNs using Keras. I am using the theano backend. I don't understand how to set values to: batch_size, steps per epoch, validation_steps. What should be the value set to ...
15
votes
2answers
10k views

High-dimensional data: What are useful techniques to know?

Due to various curses of dimensionality, the accuracy and speed of many of the common predictive techniques degrade on high dimensional data. What are some of the most useful techniques/tricks/...
15
votes
5answers
3k views

Beginner math books for Machine Learning

I'm a Computer Science engineer with no background in statistics or advanced math. I'm studying the book Python Machine Learning by Raschka and Mirjalili, but when I tried to understand the math of ...
27
votes
3answers
24k views

StandardScaler before and after splitting data

When I was reading about using StandardScaler, most of the recommendations were saying that you should use StandardScaler before ...
6
votes
2answers
13k views

Python or R for implementing machine learning algorithms for fraud detection [closed]

I was wondering which language can I use: R or Python, for my internship in fraud detection in an online banking system: I have to build machine learning algorithms (NN, etc.) that predict transaction ...
33
votes
6answers
19k views

Encoding features like month and hour as categorial or numeric?

Is it better to encode features like month and hour as factor or numeric in a machine learning model? On the one hand, I feel numeric encoding might be reasonable, because time is a forward ...
10
votes
4answers
900 views

What initial steps should I use to make sense of large data sets, and what tools should I use?

Caveat: I am a complete beginner when it comes to machine learning, but eager to learn. I have a large dataset and I'm trying to find pattern in it. There may / may not be correlation across the data,...
7
votes
1answer
624 views

data science / machine learning resources? [closed]

In a few weeks I'm starting a new job that will be involved in machine learning and data science. I have a masters degree in probability / mathematics but I have no knowledge of machine learning and ...
20
votes
2answers
5k views

local minima vs saddle points in deep learning

I heard Andrew Ng (in a video I unfortunately can't find anymore) talk about how the understanding of local minima in deep learning problems has changed in the sense that they are now regarded as less ...
14
votes
2answers
7k views

How to train model to predict events 30 minutes prior, from multi-dimensionnal timeseries

Experts in my field are capable of predicting the likelyhood an event (binary spike in yellow) 30 minutes before it occurs. Frequency here is 1 sec, this view represents a few hours worth of data, i ...
11
votes
2answers
3k views

Solving a system of equations with sparse data

I am attempting to solve a set of equations which has 40 independent variables (x1, ..., x40) and one dependent variable (y). The total number of equations (number of rows) is ~300, and I want to ...
10
votes
2answers
2k views

Machine Learning Steps

Which of the below set of steps options is the correct one when creating a predictive model? Option 1: First eliminate the most obviously bad predictors, and preprocess the remaining if needed, then ...
5
votes
1answer
30k views

How to download dynamic files created during work on Google Colab?

I have two different files and on the first, I tried to save data to file as: np.save(open(Q1_TRAINING_DATA_FILE, 'wb'), q1_data) On second file, i'm trying to ...
20
votes
2answers
15k views

Doc2Vec - How to label the paragraphs (gensim)

I am wondering how to label (tag) sentences / paragraphs / documents with doc2vec in gensim - from a practical standpoint. Do you need to have each sentence / paragraph / document with its own ...
11
votes
3answers
2k views

Statistics + Computer Science = Data Science? [closed]

i want to become a data scientist. I studied applied statistics (actuarial science), so i have a great statistical background (regression, stochastic process, time series, just for mention a few). But ...
8
votes
2answers
6k views

Why large weights are prohibited in neural networks?

Why weights with large values cause neural networks to be overfitted, and consequently we use approaches like regularization to neutralize weights with large values?

1
2 3 4 5
9