Questions tagged [artificial-intelligence]

Artificial Intelligence (AI) is a topic in computer science that deals with the study / creation of intelligent machines. Use this tag for on-topic questions that have an AI aspect.

Filter by
Sorted by
Tagged with
0
votes
0answers
7 views

Can Grad-CAM be used for CNNs with a flatten layer?

I would like to visualize CNNs with a flatten layer. I looked into Grad-CAM, which is one of the most popular visualization methods for CNNs, but I thought it could only be used for CNNs with a global ...
0
votes
0answers
10 views

Why applying Grad-CAM to the input layer is not common?

Grad-CAM is a popular tool that could be applied to the last convolutional layer to understand the inner structure of the deep neural network. It is general to apply Grad-CAM onto the last ...
0
votes
0answers
19 views

Which is the correct ROC metric type to use for Imbalanced Dataset (macro,micro, weighted) for multiclass classification?

I am working on a classification task where I have more than 350 classes with HUGE data imbalance. I want to check my model performance but I am confused which metric to use. So I Googled some ...
0
votes
0answers
24 views

How can I master the Maths behind Machine Learning / Deep Learning? [duplicate]

I am a Computer Engineer , who is interested in understanding and properly interpreting the mathematics behind the ML and DL algorithms I have worked with some classification algorithms but regression ...
2
votes
1answer
48 views

Objective functions of the flow network based generative model by Yoshua Bengio?

I am reading the Yoshua Bengio et al, Flow Network based Generative Models for Non-Iterative Diverse Candidate Generation. The objective function, Equation (11) and (12) are set for a given trajectory ...
2
votes
1answer
34 views

Rationale for the objective function in a flow network based generative model by Yoshua Bengio?

I am reading the Yoshua Bengio et al, Flow Network based Generative Models for Non-Iterative Diverse Candidate Generation. It seems to me the objective of the paper is to generate the flow $F$ given ...
0
votes
0answers
12 views

what we can use instead of mean of class in fisher’s linear discriminant?

I Need to know can we use another soulation to calculate bias that is center of mean for each class(1,2) ??? actually i need to use another way for bias in fisher’s linear discriminant instead of ...
1
vote
1answer
29 views

700,000 data points are true values while 1,300,000 data points are false values, is it an imbalanaced dataset? [closed]

I am trying to build a decision tree model and I have 700,000 true values while 1,300,000 data points are false values, in total, I have 2,000,000 data points including duplicates. I am wondering if ...
0
votes
0answers
12 views

Issue related to Transformer decoder druing Inference using all previous output tokens until each decoder time step

I've been trying to understand the shapes used during decoder (both self-attention and enc-dec-attention blocks) and understand there is a difference in the way decoder runs during training versus ...
0
votes
0answers
26 views

Gradient descent - no way to find the global optimum if the model is stuck at local optima?

When I was learning about gradient descent a few minutes ago, I looked at the equation This is supposed to find the slope of a point on the cost function J(θ0, θ0), and then go in the opposite ...
2
votes
2answers
122 views

Gradient descent - why the partial derivative?

I'm quite new to AI/ML, and I was learning about gradient descent. I saw this equation that explained the gradient descent algorithm: I quite understood everything except the reason this equation ...
0
votes
0answers
43 views

Which machine learning or AI algorithm is best for this scenario?

I have a program which takes 3 inputs (a, b, c) and produces 2 outputs (x, y). The program is deterministic; if I give the same inputs, I'm guaranteed to get the same outputs. My objective is find ...
0
votes
0answers
20 views

Models for Long-Term Time-Series Forecasting and Pattern Recognition

I'm trying to find a solution for long-term electricity hourly prices forecasting. Explaining simply, I have some data from 2018 - 2021 containing Demand, Renewable Generation, Hydropower Generation, ...
12
votes
2answers
3k views

Foundation models : Is it a new paradigm for statistics and machine learning?

A recent debate on so called Foundation models (CRFM) brings a real question of if we can build very large models on any specified domain, similar to current large language models, and replace our any ...
5
votes
1answer
204 views

Why is naive Bayes overconfident?

In the fourth edition of "Artificial Intelligence: a modern approach" by Russel and Norvig, they write in section 12.6, regarding the Naive Bayes Model for text classification, the following:...
0
votes
0answers
24 views

Which predictive model is best fit here?

I have two predictor variable both numeric with right skew. My outcome variable is binary as positive and negative. Sample size id 157 and positive cases are only 10. That's just 6.37%. I know there ...
0
votes
0answers
33 views

Are goal-reaching and optimizing the utility function special cases of performance measure?

I don't know if it's ok to ask it here but in the book Artificial Intelligence: A Modern Approach by Russell, performance measure is defined as something evaluating the behavior of the agent in an ...
0
votes
0answers
39 views

Why transformer in deep learning is called transformer?

Where does the name "transformer" come from in deep learning? I want to know more about the correlation between its name and its working principle.
0
votes
1answer
45 views

Is it possible to run Spiking Neural Network (SNN) on the current von neumann architecture?

I am a new to Spiking Neural network (SNN). I read a couple of papers about it. Some of them highlighted that SNN is a kind of a hardware-dependent model that can efficiently work on neuromorphic ...
1
vote
0answers
24 views

Understanding convolutional operators in graphs

In a Graph Convolutional Neural Network, a convolutional operation is $$ h_i^{(l)} = \sigma\left(W^{(l)} \cdot \text{Agg}\left\{ h_j^{(l-1)},\forall j \in \hat{N}(i)\right\} \right)$$ Is my intuition ...
0
votes
1answer
88 views

What is the intuition behind weights and biases in machine learning?

I am having a hard time trying to understand what weights and biases mean in a neural network and how do they help the neural network make right prediction . I know how weights and biases are changed ...
1
vote
0answers
94 views

Validation accuracy drops with ResNet50 augmented training

I am trying to train a ResNet50 from scratch on Imagenette https://github.com/fastai/imagenette I started by just directly training and had the result below: Orange: training dataset. Blue: ...
1
vote
0answers
7 views

visual of how to find the feature maps of two input with 3 filters in cnn

1st layer of CNN : We have an input and we apply two filters on the input to get 2 feature maps. now in my knowledge, we'll use these two feature maps as an input in 2nd layer of CNN. if we apply 3 ...
0
votes
0answers
13 views

Initial weights of ANN close to zero [duplicate]

My question is rather short. Why do we initialize the weights of an ANN close to zero? I understand the benefit with regard to an early stopping regularization technique but besides from that is there ...
0
votes
1answer
20 views

How to use data augmentation in the context of model evaluation in machine learning?

I'm trying to use data augmentation for training a model for a classification task. But I'm not sure about how to use data augmentation in a fair and meaningful way in the evaluation of a machine ...
1
vote
2answers
39 views

Can a neural network nodes "underweight" or "overweight" themselves?

I was under an impression that artificial intelligence is modelled after organic intelligence. Under the context of organic intelligence, it seems that some individuals are capable of getting caught ...
2
votes
1answer
85 views

Does the transformers model (in “Attention is All You Need”) exclude the encoder in language modelling tasks?

The language model I am referring to is the one outlined in "Attention is All You Need": My understanding is (please correct me if I am wrong) that when the task is translation, the encoder'...
1
vote
0answers
11 views

Normalization Techniques in Deep Learning

I understand different Normalization techniques in Deep Learning eg. BatchNorm, LayerNorm, InstanceNorm, GroupNorm, but I don't currently understand which normalization technique should be used where. ...
0
votes
0answers
18 views

Machine Learning for Product Configuration

I have a hypothetical to ask regarding Machine Learning. Let's use bicycles as an example. Let's say I assemble and sell bicycles. I source components from a variety of OEMs to create really cool ...
0
votes
0answers
7 views

Applications of Associative Memory Models

Associative memory models, such as Hopfield Networks and newer dense associative memories (Krotov) have been an important field of research in the last decades. I have read quite some papers, both new ...
0
votes
1answer
21 views

Multiple linear regression with very large cost

So, I'm trying to enter the Data Science world but struggling with a very simpel exercise. I'm using a dataset to get personal medical costs from a bunch of personal info. The database look like this: ...
0
votes
0answers
51 views

Need Accuracy and Loss Graphs Explanation

My training and validation accuracy graph are near to each other. The validation accuracy is like 0.05-0.3 more than training accuracy. And the training loss and validation loss are also near like ...
0
votes
0answers
156 views

Importance of the shape of loss curves in deep learning

There is this common assumption/perception of how a "good" loss curve should look like. The famous image below illustrates this quite well (However recent research from openAI discovered a ...
0
votes
0answers
11 views

How can I find the optimal time to update data mart?

I am assigned to find out the optimal or some reasonable time when to update a table in the data mart I am working on. The data that I need to update in the data mart are collected from air ...
0
votes
0answers
35 views

Which learning technique fits for training a Pokemon combat agent?

My goal is to train a combat agent for pokemon battles with some reinforced or deep learning techniques. I'm a software developer and don't have a lot of background for this problems but I've been ...
0
votes
0answers
16 views

Relationship between number of NN parameters and required training iterations

How does the number of parameters influence the training time in terms of iterations? I'm aware that the training takes more hours the more parameters you have, but would like to know whether the ...
0
votes
0answers
6 views

Simple Time Series problem with RNN/LSTM in Keras

I started to get into RNNs for time series analysis, coming from a statistical background. I wanted to try with a simple toy example: X = [1,2,3,4,5,6,7,8,9,10] The aim is to get as output the input ...
1
vote
1answer
21 views

Can Neurons (Features) be repeated in Dense Layers

I was revising Convolutional Neural Networks and encountered the following question. If I were to classify a cat and a dog (the famous cats vs dog classifier), then assuming there are 2 Dense Layers ...
0
votes
0answers
20 views

Need Help in Solving a Real World Data Science Problem!

I work as an Analyst in a firm and due to privacy issues, I cannot disclose the name or the industry I am working on. But the problem that I am working on is finding insights into how these variables ...
4
votes
1answer
240 views

Viterbi Algorithm - Most likely sequence vs sequence of most likely states

I'm trying to understand why the following pseudo-code function is correct: ...
2
votes
1answer
999 views

What's the pros and cons between Huber and Pseudo Huber Loss Functions?

The Huber Loss is: $$ huber = \begin{cases} \frac{1}{2} t^2 & \quad\text{if}\quad |t|\le \beta \\ \beta |t| &\quad\text{else} \end{cases} $$ The pseudo huber is: $$...
1
vote
0answers
23 views

What is the difference between Reinforment Learning and Sequential Pattern Mining? [closed]

I am kind of confused with sequential pattern mining and reinforcement learning. I just want to know the difference between sequential pattern mining and reinforcement learning. With sequential ...
0
votes
0answers
24 views

Is there any mathematical results that quantifies the error that a deep-learning network does while approximating a function?

I am working with kernel methods (AKA Support Vector Machines), and found that they outperform systematically deep-learning methods, in all the numerical experiments I have done. Indeed, I noticed ...
0
votes
0answers
15 views

Question on Machine Learning to Generate Genetic Algorithms with Highest Sensitivity and Specificity Values

I am new to this website, but I am a researcher who has no experience in ML who is trying to generate molecular algorithms that use a combination of genetic markers in order to correctly predict ...
3
votes
0answers
60 views

Clustering large set of images

I've got some big datasets of images (a few million each), and I would like to cluster them according to images' visual similarities. I've extracted a feature vector for each image; the space of ...
1
vote
1answer
19 views

Why are non-local blocks really slow?

I have added non-local blocks into my CNN decoder, but this increases the training time by four times the original time without, so i don't really want to use it! What is the reason it is too slow? My ...
0
votes
1answer
25 views

Why is Inception Net so bad for my model?

I am running Inception Encoder Net with Batch Normalisation on images of cell nuclei, in particular, there are thousands of data points to be classified into 5 groups. I have run various ResNet models ...
0
votes
0answers
36 views

How can I do one class learning for outlier detection?

I understand I can use various sampling techniques when dealing with imbalanced datasets. However, I wonder how I can build a classification model from the training dataset only including data that ...
2
votes
0answers
321 views

Are decision trees related to maximum likelihood estimation?

I am just curious if decision tree is conceptually related to maximum likelihood estimation. As far as I know, when a decision tree split, it uses entropy for its nodes. However, I don't know if using ...
1
vote
0answers
74 views

How does self play work in RL?

I try to understand how self play worl in Reinforcement learning with ppo2. Do i have one model which predicts an action for both sides? If yes, do i Update the model with experiences ...

1
2 3 4 5 6