74

Introduction The term self-supervised learning (SSL) has been used (sometimes differently) in different contexts and fields, such as representation learning [1], neural networks, robotics [2], natural language processing, and reinforcement learning. In all cases, the basic idea is to automatically generate some kind of supervisory signal to solve some task (...


14

Self-supervised learning is when you use some parts of the samples as labels for a task that requires a good degree of comprehension to be solved. I'll emphasize these two key points, before giving an example: Labels are extracted from the sample, so they can be generated automatically, with some very simple algorithm (maybe just random selection). The task ...


13

Things in italics should give you enough googleable terms to start a deeper dive :P. There are 3 main branches of statistical ML. Supervised Learning This approach is taken when a problem can be phrased as associating some $X$ with some $Y$. For example, classifying a picture of a cat ($X$) with the label “Cat” ($Y$). Training in supervised learning usually ...


9

Any supervised learning (SL) problem can be cast as an equivalent reinforcement learning (RL) one. Suppose you have the training dataset $\mathcal{D} = \{ (x_i, y_i \}_{i=1}^N$, where $x_i$ is an observation and $y_i$ the corresponding label. Then let $x_i$ be a state and let $f(x_i) = \hat{y}_i$, where $f$ is your (current) model, be an action. So, the ...


8

Semi-supervised learning Semi-supervised learning is the collection of machine learning techniques where there are two datasets: a labelled one and an unlabelled one. There are two main problems that can be solved using semi-supervised learning: transductive learning (i.e. label the given unlabelled data) and inductive learning (generalization) (i.e. find a ...


7

Note: you mentioned in the comments that you are reading the old, pre-print version of the paper describing AlphaZero on arXiv. My answer will be for the "official", peer-reviewed, more recent publication in Science (which nbro linked to in his comment). I'm not only focusing on the official version of the paper just because it is official, but also because ...


6

Both semi-supervised and self-supervised methods are similar in the sense that the goal is to learn with fewer labels per class. The way both formulate this is quite different: Self-Supervised Learning: This line of work aims to learn image representations without requiring human-annotated labels and then use those learned representations on some ...


6

The concept of "planning" is not just related to RL. In general (as the name suggests), planning consists in creating a "plan" which you will use to reach a "goal". The goal depends on the context or problem. For example, in robotics, you can use a "planning algorithm" (e.g. Dijkstra's algorithm) in order to find the ...


5

Self-supervised visual recognition is often applied to representation learning. Here we first learn features on unlabeled data (representation learning), and then learn the real model on features extracted from the labeled data. This especially makes sense when we have a lot of unlabeled data and few labeled data. The features can be learned by solving so ...


5

The way children learn is in many ways supervised. It is true that certain abilities are there genetically (visual system, object recognition, to large extent voice recognition), but a lot of human experience is gained as a result of response, either from the environment or from the mentor (parent, teacher, etc). Social interaction is possible only when a ...


5

The previous answer has given a good insight into the difference between two areas. I would like to give more examples. Semi-Supervised Learning work with improving the data set by adding up new examples. There are iterative systems where we train a model on a given dataset and improve the model further after deploying it on the real world by adding ...


5

In reinforcement learning, exploration has a specific meaning, which is in contrast with the meaning of exploitation, hence the so-called exploration-exploitation dilemma (or trade-off). You explore when you decide to visit states that you have not yet visited or to take actions you have not yet taken. On the other hand, you exploit when you decide to take ...


4

So in a sense you are correct. Using your jargon: linear regression will only "work" if the true function is approximately $y=h(x)=\beta^{T}x+\beta_0$. Advantages to using this is that its light, its convex, and all-around easy. but for alot of larger problems, this wont work. As you said you want the machine to do the work, so this is (kinda) where deeper ...


4

Imitation learning is supervised learning applied to the RL setting. In any general RL algorithm (such as Q-learning), the learning is done on the basis of the reward function. However, consider a scenario where you have available the optimal policy in the form of a table, mapping each state to each action. In this scenario you will not care about the ...


3

If you learn a policy or a value function from experience (that is, interaction with an environment), that's RL. In the case of AlphaGo, the MCTS is used to acquire the experience. RL could in fact be considered supervised learning (SL) or, more specifically, self-supervised learning, where the experience corresponds to the labels in SL, especially nowadays ...


3

Automated machine learning (AutoML) is an umbrella term that encompasses a collection of techniques (such as hyper-parameter optimization or automated feature engineering) to automate the design and application of machine learning algorithms and models. Reinforcement learning (RL) is a sub-field of machine learning concerned with the task of making ...


3

Decision Tree learners, on their own, are not a good way to deal with imbalanced data. The most commonly used algorithms, by default, make no attempt to address this problem. If you look carefully at the three sources you post, you will find that they actually all agree on this point. Two of the sources actually propose methods of addressing this ...


3

This formulation/interpretation can indeed be confusing (or even misleading), as the output of a neural network is usually deterministic (i.e. given the same input $x$, the output is always the same, so there is no sampling), and there isn't really a probability distribution that models any uncertainty associated with the parameters of the network or the ...


3

The programmer already guides the RL algorithm (or agent) by specifying the reward function. However, the reward function alone may not be sufficient to learn efficiently and fast, as you correctly noticed. To attempt to solve this inefficiency problem, one solution is to combine reinforcement learning with supervised learning. For example, the paper Deep Q-...


3

Supervised learning The supervised learning (SL) problem is formulated as follows. You are given a dataset $\mathcal{D} = \{(x_i, y_i)_{i=1}^N$, which is assumed to be drawn i.i.d. from an unknown joint probability distribution $p(x, y)$, where $x_i$ represents the $i$th input and $y_i$ is the corresponding label. You choose a loss function $\mathcal{L}: ...


3

In Supervised learning, the goal is to learn a mapping from points in a feature space to labels. So that for any new input data point, we are able to predict its label. whereas in Unsupervised learning data set is composed only of points in a feature space, i.e. there are no labels & here the goal is to learn some inner structure or organization in the ...


3

Reward in reinforcement learning (RL) is entirely different from a supervised learning (SL) label, but can be related to it indirectly. In a RL control setting, you can imagine that you had a data oracle that gave you SL training example and label pairs $x_i, y_i$ where $x_i$ represents a state and $y_i$ represents the correct action to take in that state in ...


3

In a nutshell : Memorizing is not Learning So, first let's just remind the classical use of a neural net, in Supervised Learning : You have a set of $(x_{train}, y_{train}) \in X \times Y$ pairs, and you want to extract a general mapping law from $X$ to $Y$ You use a neural net function $f_{\theta} : x \rightarrow f_{\theta}(x)$, with $\theta$ the weights (...


2

The GA will require a fitness function, which means you need labeled data for comparison. That conclusion is wrong. Yes, sometimes your fitness function will use labeled data. For example, if you want to train an XOR gate or any other known function. However, there is arguably no advantage of training a function with neuroevolution versus backpropagation, ...


2

You want to compute the mean loss over all batches. What you need to do is to divide the sum of batch losses with the number of batches! In your case: You have a training set of $21700$ samples and a batch size of $500$. This means that you take $21700/500 \approx 43$ training iterations. This means that for each epoch the model is updated $43$ times! So ...


2

Accuracy can sometimes be a very coarse metric. When it is applied to three class problems, people often take the class label with maximum predicted probability and predict that. The probabilities of the individual labels are ignored. I'd recommend that as well as accuracy you calculate sensitivity and specificity for each class and the area under the ROC ...


2

To know if your model needs more training data, try to plot out "learning curves", that are based on increasing size of the training set. Basically, you calculate training and validation accuracy metrics for 1, 2, 3, 4, 5, ..., m training samples. Size of validation set may be constant over time. If the accuracy is still rising when your data set is fully ...


2

By "immediate vector-valued feedback", they probably mean exactly the label in the "labeled examples" you mentioned.


2

RL can be used in the context of Neural Architecture Search (NAS), with is a form of automated ML. A model searches for an architecture that performs a given task. How well this task is performed guides how the architecture will be modified (improved) on the next pass. It works but is very computation-intensive (think hundreds of GPUs). See for instance: B....


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