Questions tagged [theory]

Use for questions on AI theory and other theoretical subjects related to AI.

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

The convolutional network architectures with enhanced invariance

It is well known, that CNN have advantage with respect to the Dense neural networks in the image classification and other pattern recognition tasks, because they have a translationall invariance built ...
1
vote
1answer
31 views

The are some fundamental learning theories for developing an AI that imitates human behavior

Most if not all AI systems are not to imitate human, but to finally out-perform human. Examples include using AI to play a game, classification problems, auto-driving, and goal-oriented chatbots. ...
0
votes
0answers
33 views

Is it a good idea to train a neural network to classify images without base-hypothesis?

I'm a relative beginner in deep-learning (understand by that, I'm doing my first kaggle competition right now, and I have loads to learn still) and I was just wondering something. Let's say you have ...
1
vote
1answer
57 views

How artificial intelligence will change the future?

AI is the emerging field and biggest business opportunity of the next decade. It's already automating manual and repetitive tasks. And in some areas, it can learn faster than humans, if not yet as ...
0
votes
1answer
86 views

What is the relationship between gradient accumulation and batch size?

I am currently training some models using gradient accumulation since the model batches do not fit in GPU memory. Since I am using gradient accumulation, I had to tweak the training configuration a ...
2
votes
0answers
37 views

Is there any application of topology to deep learning?

Is there any application of topology (as in math discipline) to deep learning? If so, what are some examples?
0
votes
2answers
37 views

Is it theoretically possible (or impossible) that principal component analysis worsens the performance of the model?

In case I had a prediction model and decided to add a PCA step prior to the model, is it theoretically possible/impossible that the number of output dimensions that is better for all tests may perform ...
2
votes
0answers
32 views

Is there a theory that captures the following ideas?

A big class of problems that are relevant in today's society are full of uncertainty and are also sometimes computationally intractable. Along our lives we come to realize that we are solving the same ...
1
vote
0answers
19 views

Are there any good resources about techniques used for entity extraction

Given some natural language sentences like "I would like to talk to Mr. Smith" I would like to extract entities like the person "Smith". I know that frameworks exist which are capable of doing so (f. ...
1
vote
2answers
72 views

What are some examples of classical AI methods?

All of the attention in the AI field over the past few years has been toward statistical AI, now that Machine Learning has been validated on hard problems. However, I want to learn more about ...
1
vote
1answer
41 views

What is the theoretical basis for the use of Cross Validation set?

So let's follow this line of reasoning. We use a MLE estimator (implementation doesn't matter) and we have a train set. We assume that we have sampled training set from a Gaussian distribution $\...
2
votes
0answers
18 views

Suitable kernels for Gaussian processes

Consider a stochastic process $\{X_t \colon t \in T\}$ indexed by a set $T$. We assume for simplicty that $T \in \mathbb{R}^n$. We assume that for any choice of indexes $t_1, \dots, t_n$, the random ...
0
votes
2answers
111 views

Understanding CNN in a few sentences

I don't know if this is the right place to ask this question. If it is not, please tell me and I remove it. I've just started to learn CNN and I'm trying to understand what they do and how they do it....
4
votes
1answer
63 views

Is there a way to ensure that my model is able to recognize an unseen example?

My question is more theoretical than practical. Let's say that I am training my cat classifier with a dataset that I feel is pretty representative of cat images in general. But then a new breed of cat ...
5
votes
5answers
281 views

What event would confirm that we have reached AGI or a highly intelligent system?

I was listening to a podcast on the topic of AGI and a guest made an argument that if strong music generation were to happen, it would be a sign of true intelligence in machines because of how much ...
2
votes
0answers
26 views

Did people analyze dynamics of very simple LSTMs?

I wonder if researchers tried to understand how LSTMs work by analyzing the dynamics of simple LSTM (e.g. with 2 units)? For example how the hidden state evolves depending on the properties of weight ...
2
votes
1answer
83 views

What make a CNN suitable for image classification or for semantic segmentation?

I've just started with CNN and there is something that I haven't understood yet: How do you "ask" a network: "classify me these images" or "do semantic segmentation"? I think it must be something on ...
2
votes
0answers
126 views

What does “episodic training” mean?

I'm reading the book Hands-On Meta Learning with Python, and in Prototypical networks said: So, we use episodic training—for each episode, we randomly sample a few data points from each class in ...
2
votes
1answer
47 views

How many ways are there to perform image segmentation?

I'm new in Artificial Intelligence and I want to do image segmentation. Searching I have found these ways Digital image processing (I have read it in this book: Digital Image Processing, 4th edition)...
2
votes
0answers
75 views

Is this a problem well suited for machine learning?

The light-field of a certain scene is the set of all light rays that travel through the volume of that scene at a specific point of time. A light-field camera for example captures and stores a subset ...
4
votes
2answers
166 views

Are current AI models sufficient to achieve Artificial General Intelligence?

I read an interesting essay about how far we are from AGI. There were quite a few solid points that made me re-visit the foundation of AI today. A few interesting concepts arose: imagine that you ...
1
vote
0answers
31 views

How would you differentiate between different on-policy reinforcement learning algorithms?

How would you differentiate between different on-policy reinforcement learning algorithms?
0
votes
1answer
78 views

When are compiled vs. interpreted languages more optimal in AI?

When are interpreted languages more optimal? When are compiled languages more optimal? What are the qualities and functions that render the so in relation to various AI methods?
2
votes
1answer
57 views

Is there anything theoretically revolutionary about Deep Neural Network?

In recent years we have seen quite a lot of impressive display of Deep Neural Network (DNN), as demonstrated most famously by AlphaGo and its cousin programs. But if I understand correctly, deep ...
2
votes
1answer
875 views

What is convergence in machine learning?

I came across this answer on Quora, but it was pretty sparse. I'm looking for specific meanings in the context of machine learning, but also mathematical and economic notions of the term in general.
3
votes
1answer
681 views

What does 'democratizing AI' exactly mean?

In my AI literature research, I often notice authors use term 'democratizing AI', especially in the AutoML area. I think I have an idea of what this means, but I would like to ask you for some more ...
11
votes
3answers
150 views

Is there a way to understand neural networks without using the concept of brain?

Is there a way to understand, for instance, a multi-layered perceptron without hand-waving about them being similar to brains etc? For example: it is obvious that what a perceptron does is ...
1
vote
1answer
39 views

Why is image classification tasks are dominated by minimizing cost function instead of maximizing ones?

I was watching a video of policy gradient by Andrej Karpathy at 10:00 there shows an equation for supervised learning for image classification. $max\sum _{i}log \:p(y_i|x_i)$ I have worked with ...
5
votes
2answers
53 views

Is the gradient at a layer independent of the activations of the previous layers?

Is the gradient at a layer (of a feed-forward neural network) independent of the activations of the previous layers? I read this in a paper titled Mean Field Residual Networks: On the Edge of Chaos (...
4
votes
0answers
51 views

Is it possible to control asymptotic behaviour of neural network models?

Is it possible to specify what the asymptotic behaviour of a Neural Networks (NN) model should be? I am thinking on NN which try to learn a mapping $\vec y=f(\vec x)$ with $\vec x$ a vector of ...
6
votes
9answers
818 views

Can Neural Networks self-optimize?

Suppose that you show a neural network its own code, and allow it to edit itself? Can a neural network modify its own weights and architecture (the number of layers, the number of neurons per layer, ...
47
votes
4answers
11k views

Are neural networks prone to catastrophic forgetting?

Imagine you show a neural network a picture of a lion 100 times and label with "dangerous", so it learns that lions are dangerous. Now imagine that previously you have shown it millions of images of ...
2
votes
0answers
43 views

Why do neural networks have bias units?

Why do neural networks have bias units? Why is it sometimes okay to opt them out?
1
vote
1answer
1k views

What is an agent in Artificial Intelligence?

While studying artificial intelligence, I have often encountered the term "agent" (often autonomous, intelligent). For instance, in fields such as Reinforcement Learning, Multi-Agent Systems, Game ...
2
votes
1answer
55 views

Is explainable AI more feasible through symbolic AI or soft computing?

Is explainable AI more feasible through symbolic AI or soft computing? How much each paradigm, symbolic AI and soft computing (or hydrid approaches), adresses explanation and argumentation, where ...
2
votes
2answers
133 views

Would an artificial general intelligence have to be Turing complete?

Artificial general intelligence is the intelligence of a machine that could successfully perform any intellectual task that a human being can. Would an artificial general intelligence have to be ...
2
votes
1answer
62 views

What are the known ways to test AI black boxes?

Is there any conmprehensive paper or site that describes an overview of the known black box testing techniques?
1
vote
0answers
17 views

Theoretical grounding for ease of training with a prior

If we have a neural network that learns the generative model for P(A, B, C) And now, we want to learn the generative model for P(A, B, C, D) Is there any theory that says learning P(A,B,C) and then ...
0
votes
0answers
25 views

Superhuman AI to disguise as human players

TL;DR below. You entered a online mini video game. It's a fast-paced hypercasual game. But it's a ranked one - top 10 get prizes. The leaderboard resets and prizes are handed out every couple of ...
2
votes
1answer
95 views

Neural nets need a short term memory of some sort

I notice with the recent revelation of severe limitations in some AI domains such as self driving cars that NNets behave with the same sort of errors as in simpler models. Ie: They may be ~100% ...
1
vote
1answer
37 views

Is Intelligence a naturally occurring function of Information Technology?

Here intelligence is defined as any analytic or decision making process, regardless of strength (utility), and, potentially, any process of computation that produces output, regardless of the medium. ...
1
vote
3answers
68 views

Are winning and imitating the primary indicators of intelligence?

Conjecture 1: The smartest chess playing system is the one that wins the tournament. Conjecture 2: The computer system that imitates human dialog is as smart as the human whose dialog was imitated. ...
3
votes
1answer
87 views

How would a quantum computer potentially facilitate artificial consciousness, assuming it is possible?

How would a quantum computer potentially facilitate artificial consciousness, assuming it is possible?
3
votes
0answers
27 views

Batch PTA stopping condition

I am reviewing my Neural Network lectures and I have a doubt: My book's (Haykin) batch PTA describes a cost function which is defined over the set of the misclassified inputs. I have always been ...
2
votes
0answers
70 views

An idea I had to create the first Humanoid using Deep Learning

I've come up with an idea on how we could use a combination of Deep Learning and body sensors to create a walking talking living humanoid. Here goes: First, we will recruit 1 billion people and have ...
7
votes
2answers
124 views

What does learning mean?

Can someone explain what is the process of learning? What does it mean to learn something?
3
votes
0answers
199 views

Are human brain processes, like creativity, intuition or imagination, computable processes?

Are human brain processes, like intuition, creativity, imagination and the ability to create art, computable processes?
0
votes
1answer
84 views

How can I systematically learn about the theory of neural networks?

I have seen a few articles about neural nets. Mostly they went along these lines: we tried these architectures, these meta parameters, we trained it for $x$ hours on $y$ CPUs, and it gave us these ...
2
votes
1answer
143 views

Can NEAT produce neural networks where inputs are directly connected to outputs?

Can NEAT produce neural networks where inputs are directly (without intermediate hidden neurons) connected to outputs?
3
votes
2answers
115 views

Will a neural network always predict the correct label if it sees the exact same input during training and testing?

If I'm performing a text classification task using a model built in Keras, and for example, am attempting to predict the appropriate tag given a Stack Overflow question: "how to subtract 1 from an ...