Apprentissage supervisé | EPFL Graph Search

L'apprentissage supervisé (supervised learning en anglais) est une tâche d'apprentissage automatique consistant à apprendre une fonction de prédiction à partir d'exemples annotés, au contraire de l'apprentissage non supervisé. On distingue les problèmes de régression des problèmes de classement. Ainsi, on considère que les problèmes de prédiction d'une variable quantitative sont des problèmes de régression tandis que les problèmes de prédiction d'une variable qualitative sont des problèmes de classification. Les exemples annotés constituent une base d'apprentissage, et la fonction de prédiction apprise peut aussi être appelée « hypothèse » ou « modèle ». On suppose cette base d'apprentissage représentative d'une population d'échantillons plus large et le but des méthodes d'apprentissage supervisé est de bien généraliser, c'est-à-dire d'apprendre une fonction qui fasse des prédictions correctes sur des données non présentes dans l'ensemble d'apprentissage. Définition mathématiq

Learning Long-Range Perception Using Self-Supervision From Short-Range Sensors and Odometry

Mirko Nava

We introduce a general self-supervised approach to predict the future outputs of a short-range sensor (such as a proximity sensor) given the current outputs of a long-range sensor (such as a camera). We assume that the former is directly related to some piece of information to be perceived (such as the presence of an obstacle in a given position), whereas the latter is information rich but hard to interpret directly. We instantiate and implement the approach on a small mobile robot to detect obstacles at various distances using the video stream of the robot's forward-pointing camera, by training a convolutional neural network on automatically-acquired datasets. We quantitatively evaluate the quality of the predictions on unseen scenarios, qualitatively evaluate robustness to different operating conditions, and demonstrate usage as the sole input of an obstacle-avoidance controller. We additionally instantiate the approach on a different simulated scenario with complementary characteristics, to exemplify the generality of our contribution.

2019

Theory of representation learning in cortical neural networks

Carlos Stein Naves de Brito

Our brain continuously self-organizes to construct and maintain an internal representation of the world based on the information arriving through sensory stimuli. Remarkably, cortical areas related to different sensory modalities appear to share the same functional unit, the neuron, and develop through the same learning mechanism, synaptic plasticity. It motivates the conjecture of a unifying theory to explain cortical representational learning across sensory modalities. In this thesis we present theories and computational models of learning and optimization in neural networks, postulating functional properties of synaptic plasticity that support the apparent universal learning capacity of cortical networks. In the past decades, a variety of theories and models have been proposed to describe receptive field formation in sensory areas. They include normative models such as sparse coding, and bottom-up models such as spike-timing dependent plasticity. We bring together candidate explanations by demonstrating that in fact a single principle is sufficient to explain receptive field development. First, we show that many representative models of sensory development are in fact implementing variations of a common principle: nonlinear Hebbian learning. Second, we reveal that nonlinear Hebbian learning is sufficient for receptive field formation through sensory inputs. A surprising result is that our findings are independent of specific details, and allow for robust predictions of the learned receptive fields. Thus nonlinear Hebbian learning and natural statistics can account for many aspects of receptive field formation across models and sensory modalities. The Hebbian learning theory substantiates that synaptic plasticity can be interpreted as an optimization procedure, implementing stochastic gradient descent. In stochastic gradient descent inputs arrive sequentially, as in sensory streams. However, individual data samples have very little information about the correct learning signal, and it becomes a fundamental problem to know how many samples are required for reliable synaptic changes. Through estimation theory, we develop a novel adaptive learning rate model, that adapts the magnitude of synaptic changes based on the statistics of the learning signal, enabling an optimal use of data samples. Our model has a simple implementation and demonstrates improved learning speed, making this a promising candidate for large artificial neural network applications. The model also makes predictions on how cortical plasticity may modulate synaptic plasticity for optimal learning. The optimal sampling size for reliable learning allows us to estimate optimal learning times for a given model. We apply this theory to derive analytical bounds on times for the optimization of synaptic connections. First, we show this optimization problem to have exponentially many saddle-nodes, which lead to small gradients and slow learning. Second, we show that the number of input synapses to a neuron modulates the magnitude of the initial gradient, determining the duration of learning. Our final result reveals that the learning duration increases supra-linearly with the number of synapses, suggesting an effective limit on synaptic connections and receptive field sizes in developing neural networks.

EPFL2016

Self-supervised learning for time series classification

Anne-Sophie Brigitte Van De Velde

Time series classification (TSC) is an important and challenging problem in machine learning. In this work, we tackle the problem of TSC by first applying a Bidirectional Encoder Representations from Transformers (BERT) model, and then applying a convolutional neural network (CNN) for the classification. We report suboptimal results compared to the state-of-the-art, as the model is overfitting early in the training pro- cess. We believe that this issue might be overcome by tuning the hyperparameters more carefully.

2020

Theory of representation learning in cortical neural networks

Carlos Stein Naves de Brito

EPFL2016