# Tag Info

10

A neuron is said activated when its output is more than a threshold, generally 0. For examples : $$y = Relu(a) > 0$$ when $$a = w^Tx+b > 0$$ Same goes for sigmoid or other activation functions.

8

Supervised learning is typically an attempt to learn a mathematical function, $f(\bf X)=\bf y$. For this, you need both the input vector $\bf X$ and the output vector $\bf y$. The model outputs have whatever dimensionality that the target values have. Unsupervised learning models instead learn a structure from the data. A clustering model, for example, is ...

6

Backpropagation is a subroutine often used when training Artificial Neural Networks with a Gradient Descent learning algorithm. Gradient Descent requires computation of the error gradient, i.e. derivatives, of a cost function with respect to the network parameters. BP allows you to find this gradient a lot faster than using naive methods. Reinforcement ...

5

Well, you are definitely mixing two different things. Here are those bits: The function that deep learning approximates is basically a function that best fits the INPUT DATA points. You should not think about its differentiability or optimization aspects. We don't care what type of function it is; we just want the best fit of input data (ofcourse ...

5

As it can be easily pointed out that true random numbers cannot be generated fully by programming and some random seed is required. This is true. In fact, it is impossible to solve using software. No software-only technique can generate randomness without an initial random seed or support from hardware. This is also true for AI software. No AI design that ...

4

The term "activated" is mostly used when talking about activation functions which only outputs a value (except 0) when the input to the activation function is greater than a certain treshold. Especially when discussing ReLU the term "activated" may be used. ReLU will be "activated" when it's output is greater than 0, which is also when it's input is greater ...

4

In many cases, a production-ready model has everything it needs to make predictions without retaining training data. For example: a linear model might only need the coefficients, a decision tree just needs rules/splits, and a neural network needs architecture and weights. The training data isn't required as all the information needed to make a prediction is ...

4

The term you are looking for is stylometry, which is related to a technique in forensic linguistics called writeprint analysis. There are many different techniques to perform stylometric analysis, from the very basic 5-feature analysis classifying features such as the lexicon and idiosyncrasies unique to a person to more complex analysis utilizing neural ...

4

Trying to address all the questions asked in the end in the same order Most definitely possible. I would say its best you approach this with segmentation to start with. Just use a free GPU runtime notebook service such as Google Colab or Kaggle Kernels. But you would not directly be able to integrate with the device, you'd have to keep moving input and ...

3

Welcome to AI-SE! Please use shorter description next time because it makes it hard to understand what you ask :) I believe you want a neural network that can predict future values of multiple variables given multiple inputs. This belongs to the general time series forecasting problem. One of the best neural network architectures that can handle this ...

3

A simplex reflex agent takes actions based on current situational experiences. For example, if you set your smart bulb to turn on at some given time, let's say at 9 pm, the bulb won't recognize how the time is longer simply because that's the rule defined it follows. A simple reflex agent doesn't compute complex computational problems nor exhibit ...

3

This is too broad a topic to answer directly. If you are at the beginner stage with neural networks, you will need to learn some basic theory of the maths of neural networks, before the code will make sense. Although it is possible to write neural network code with only a vague understanding of what is going on, it is not a great way to learn for the ...

3

First of all, sigmoid does not output 0 or 1, it outputs any real number in the range between 0 and 1. Furthermore, normal neural networks doesn't output binary values, unless the output layer uses the step function as activation function (which is rare). I'm not really sure if you want the NN to be a classifier or regressor, but it sounds like you want a ...

3

The question, "Why do we need artificial intelligence?" is quite to the point. Technically, the answer is, "No reason." If we needed artificial intelligence, then we would have become extinct over the last 50,000 or more years we've been a species with human intelligence, so we want it. We do want it, and there are benefits. Some claim there are risks ...

3

The acclaimed book Artificial Intelligence: A Modern Approach (by Stuart Russell and Peter Norvig) gives a definition of an agent An agent is anything that can be viewed as perceiving its environment through sensors and acting upon that environment through actuators. This definition is illustrated by the following figure This definition (and ...

3

Hi and welcome to the community. It's important to understand these basic concepts very clearly. You have to first understand the basic unit of a neural network, a single node/neuron/perceptron. Let us forget all about Neural Networks for a bit, and talk about something far simpler. Linear Regression In the above figure, we clearly have one independent ...

2

Yes, you can train a NN to detect only one type of object like a table. However, you probably will not want to train such a NN from scratch by showing some examples of tables and non-tables. You will need to use transfer learning on a model already trained on several image classes and teach it to also recognize your new class. This transfer learning requires ...

2

Well simply put AI model can be seen just as a flowchart showing how the control flow moves where it moves how it moves why it moves etc. However AI architecture refers to the next step after building an AI model AI architecture involves representation of the functions that you use in your program. It also involved declaration of the variables you're going ...

2

Neural networks are all about taking raw input data (RGB values and pixel location) and learning useful features that are relevant to some downstream task. This process of aggregating raw inputs into higher-level features can start at the first layer past the inputs. So yes, only the first layer of the network is using the actual raw color information from ...

2

These bots depend on heavy NLP services that are provided via Azure. Implementing/deploying your own production-quality versions of these is nontrivial, if not impossible (since you don't have all the training data, internal algorithms, etc. that Microsoft/Amazon/Google/IBM et al have), and is generally non-feasible on home-grade devices. And yes, that ...

2

Architecture describes a general approach to a ML problem, and the parameterization of that approach. For example, a neural net architecture would define the number and size of different layers, the type of each layer, and so on. A model is one specific instance of a given architecture, trained on a given dataset. For the example of neural nets, the model ...

2

AI is a two step process: use data to learn a model, and then use the model to make predictions using new data. So an AI model is the result of the learning process, and the architecture is the detail of how the learning is achieved.

2

It mostly seems to be a personal preference type of thing. But in my readings, AI architecture typically means a large scale structural difference (connectionist / GOFAI; deep stack / recurrent, while AI models are finer distinctions between methods in a common architecture (say, the AlexNet vs other CNNs)

2

Since you are A C# developer already Just getting started and not sure where to go next I would suggest trying the Emotion API which is now part of the general Face API. This has the benefit of being pre-trained on a very large dataset. You can perform 30,000 recognitions/month for free.

2

A good, recent, and accessible book which includes many case studies is Prediction Machines. Check it out for more details than I can provide in this answer. Example applications are all around us, but one of the problems with recognizing them is that the bar for what we call AI is constantly being raised. Consider that a few decades ago, directions from ...

2

Such a great question. I would concur with Dennis Soemers comment that humans are not great at thinking of random numbers (just think about any card trick). However, we are very good at creating randomness through our actions. If you consider moving a computer mouse, the stock market, or playing a lottery, humans are very good at creating randomness through ...

2

Just thought I'd give an answer myself; covering the 3rd point. What are the other ways that the connections can be adjusted? What alternatives are there to using target values? I took a look on YouTube, and found https://www.youtube.com/watch?v=VnwjxityDLQ It's a only 5 minute video. The videos I was getting when I searched "Andrew​ Ng's ML course"...

2

For a foundation, there is nothing better than Cybernetics by Norbert Wiener. It is surprising how advanced this MIT professor was, prior to Turing's thought experiment on a general purpose computing machine or the embodiment of the von Neumann architecture upon which most contemporary computers are based. In key ways his analysis of time series and ...

2

If your search tree's branching factor is finite but its depth is unbounded, then DFS isn't "complete". What that means is that it may not ever find the goal node even if it exists. You would use these techniques precisely when this situation arises. To understand this, let's say you're at the node A in the following tree: A / \ / \ B C ...

2

Think of a computer as a Turing Machine--this idea is a model of computation, and all of modern computing is based on the Turing-Church thesis. Machine and program can be interchangeable—at the end of the day, it's all algorithms, whether hard coded in the form of microchip, or in the form of software. (Any microchip can be emulated as software.) Pre-...

Only top voted, non community-wiki answers of a minimum length are eligible