Questions tagged [multilayer-perceptron]

For question about Multi Layer Perceptron model/architecture, its training and other related details and parameters associated with the model.

25 questions with no upvoted or accepted answers
Filter by
Sorted by
Tagged with
9
votes
0answers
113 views

What is the number of neurons required to approximate a polynomial of degree n?

I learned about the universal approximation theorem from this guide. It states that a network even with a single hidden layer can approximate any function within some bound, given a sufficient number ...
4
votes
1answer
96 views

Backpropagation equation for a variant on the usual Linear Neuron architecture

Recently I encountered a variant on the normal linear neural layer architecture: Instead of $Z = XW + B$, we now have $Z = (X-A)W + B$. So we have a 'pre-bias' $A$ that affects the activation of the ...
4
votes
1answer
108 views

How to deal with padded inputs in a fully connected feed forward network?

I have a fully connected network that takes in a variable length input padded with 0. However the network doesn't seem to be learning and I am guessing that the high number of zeros in the input ...
3
votes
2answers
103 views

Which online machine learning technique to use for multi-class classification problem with multiple inputs?

I have the following problem. We have $4$ separate discrete inputs, which can take any integer value between $-63$ and $63$. The output is also supposed to be a discrete value between $-63$ and $63$. ...
3
votes
2answers
322 views

Unable to overfit using MLP

I'm building a 5-class classifier with a private dataset. Each data sample has 67 features and there are about 40000 samples. Samples of a particular class were duplicated to overcome class imbalance ...
2
votes
0answers
44 views

Recent algorithms for correcting mislabeled data using multilayer perceptrons

I am doing literature research on algorithms for correcting mislabeled data using multilayer perceptrons. Found an "old" paper An algorithm for correcting mislabeled data (2001) by Xinchuan Zeng et al....
2
votes
0answers
164 views

How I can predict the next number in a sequence with a neural network?

I've been dabbling with machine learning and neural networks (namely, resnet50) for a few months now, mostly doing image recognition. I am currently trying to make a program that, given a string of ...
2
votes
1answer
126 views

Is there a mathematical theory behind why MLP can classify handwritten digits?

I'm trying to really understand how multi-layer perceptrons work. I want to prove mathematically that MLP's can classify handwritten digits. The only thing I really have is that each perceptron can ...
1
vote
0answers
27 views

How can I adapt this script (to test the robustness of a perceptron) to test the robustness of a multi-layer perceptron?

The following script is from Trappenberg's Fundamentals of Computational Neuroscience and is used to test a perceptron's robustness against noise. However, how would one alter it to test the output ...
1
vote
0answers
26 views

Model unfit for some part of spiral data despite low error

I'm current testing a model for spiral data. After 500 epoches, loss is 0.04 but the result is still unmatch with some part of the training data. (bottom left) The model has 2 hidden tanh x 16 units ...
1
vote
0answers
29 views

Does a varying ANN model accuracy mean underfitting or overfitting?

Background: This is for a simulated robot with four legs, walking on a flat terrain. The ANN (an MLP) is given inputs as the robot's body angle, positions and angle of each leg with respect to the ...
1
vote
0answers
26 views

Applying Machine Learning to 2D Laser Scanner Data

We are using 2D Laser Scanner to scan various objects of different geometric shapes for e.g. cylinder, spiked, cylinder with notch, cylinder with curved edges e.t.c. The dataset contains points in the ...
1
vote
0answers
46 views

Is there data available about successful neural network architectures?

I am curious to if there is data available for MLP architectures in use today, their initial architecture, the steps that were taken to improve the architecture to an acceptable state and what the ...
1
vote
0answers
68 views

How can an ANN efficiently predict multiple numbers with fixed sum (in other words, proportions)?

I need a neural network (or any other solution) to predict 3 values which sum equals a fixed number (100). This will help me calculate proportions. Which is the most efficient way to do this? The ...
1
vote
0answers
213 views

Sigmoid output layer and Cross-Entropy cost function

I use Sigmoid activation function for neurons at output layer of my Multi-Layer Perceptron also, I use cross-entropy cost function. As I know when activation functions like Tanh is used in output ...
0
votes
0answers
40 views

What are the math theorems regarding the Multilayer Perceptron?

I've come across a theorem "Convergence theorem Simple Perceptron" for the first time, here-> https://zaguan.unizar.es/record/69205/files/TAZ-TFG-2018-148.pdf, page 27, (is in Spanish) ...
0
votes
0answers
104 views

What is the difference between the forward pass of the Multi-Layer Perceptron, Deep AutoEncoder and Deep Belief Network?

Multi-Layer Perceptron (MLP), Deep AutoEncoder (DAE), and Deep Belief Network (DBN) are trained differently. However, do they follow the same process during the inference phase, i.e., do they ...
0
votes
0answers
31 views

How to draw a 3-dimensonal shape's neural network

I am reading an exam question about NN (that I cannot publish, for copyright reasons). The question says: 'Construct a rectangle in 2D space. Define the lines, and then define the weights and ...
0
votes
0answers
14 views

NN for multivariate function interpolation

I have a multivariate function. It takes 4, real valued inputs: a, b, c, d and returns 1 complex number, z. I wanted to use Neural Networks to predict the value z' for a generic input a', b', c', d'. ...
0
votes
0answers
18 views

How to perform data augmentation on multiple input classification task?

I would like to add some more samples to my dataset which consists of two parts: 1. image and 2. numerical data. For each image in the dataset there is its corresponding numerical data as well. If it ...
0
votes
0answers
15 views

Why does 0.8:0.2 divided dataset have a much greater AUROC than its 5-fold cross validated counterpart's mean AUROC?

I trained a dataset with a 5-fold cross validation to search for hyper-parameters by an AUROC metric. For splitting I used ...
0
votes
0answers
22 views

Random Initializations with ReLU gives puzzling results

this may sound naive, but I’m getting a really puzzling result. I was experimenting with MNIST on vanilla MLP (784, 256, 128, 10) with ...
0
votes
0answers
33 views

What is the purpose to have fully connected layers?

What is the purpose of a fully connected multi layer perceptron in which every input is connected to every output by a weight? After all, the information is only distributed over several channels, but ...
0
votes
0answers
80 views

Backpropagation implementation with Java

I've been trying to implement a Multilayer Perceptron Network using java language with the ultimate goal of creating and teaching a neural network to recognize handwritten digits. Pretty simple and ...
0
votes
0answers
34 views

Problem fitting data using mlpregressor

I'm training a sklearn.neural_network.mlpregressor by a large data of students performance (an excel file with 740 students and 27 columns that are their qualities) and I want to predict their grades. ...