380 likes | 577 Views
Intelligent Environments. Computer Science and Engineering University of Texas at Arlington. Prediction for Intelligent Environments. Motivation Techniques Issues. Motivation.
E N D
Intelligent Environments Computer Science and Engineering University of Texas at Arlington Intelligent Environments
Prediction forIntelligent Environments • Motivation • Techniques • Issues Intelligent Environments
Motivation • An intelligent environment acquires and applies knowledge about you and your surroundings in order to improve your experience. • “acquires” prediction • “applies” decision making Intelligent Environments
What to Predict • Inhabitant behavior • Location • Task • Action • Environment behavior • Modeling devices • Interactions Intelligent Environments
Example • Where will Bob go next? • Locationt+1 = f(…) • Independent variables • Locationt, Locationt-1, … • Time, date, day of the week • Sensor data • Context • Bob’s task Intelligent Environments
Example (cont.) Intelligent Environments
Example • Learned pattern • If Day = Monday…Friday & Time > 0600 & Time < 0700 & Locationt = Bedroom Then Locationt+1 = Bathroom Intelligent Environments
Prediction Techniques • Regression • Neural network • Nearest neighbor • Bayesian classifier • Decision tree induction • Others Intelligent Environments
Linear Regression Intelligent Environments
Multiple Regression • n independent variables • Find bi • System of n equations and n unknowns Intelligent Environments
Regression • Pros • Fast, analytical solution • Confidence intervals • y = a ± b with C% confidence • Piecewise linear and nonlinear regression • Cons • Must choose model beforehand • Linear, quadratic, … • Numeric variables Intelligent Environments
Neural Networks Intelligent Environments
Neural Networks • 10-105 synapses per neuron • Synapses propagate electrochemical signals • Number, placement and strength of connections changes over time (learning?) • Massively parallel Intelligent Environments
Computer vs. Human Brain Intelligent Environments
Computer vs. Human Brain “The Age of Spiritual Machines,” Kurzweil. Intelligent Environments
Artificial Neuron Intelligent Environments
Artificial Neuron • Activation functions Intelligent Environments
Perceptron Intelligent Environments
Perceptron Learning Intelligent Environments
Perceptron • Learns only linearly-separable functions Intelligent Environments
Sigmoid Unit Intelligent Environments
Multilayer Network ofSigmoid Units Intelligent Environments
Error Back-Propagation • Errors at output layer propagated back to hidden layers • Error proportional to link weights and activation • Gradient descent in weight space Intelligent Environments
NN for Face Recognition 90% accurate learning head pose for 20 different people. Intelligent Environments
Neural Networks • Pros • General purpose learner • Fast prediction • Cons • Best for numeric inputs • Slow training • Local optima Intelligent Environments
Nearest Neighbor • Just store training data (xi,f(xi)) • Given query xq, estimate using nearest neighbor xk: f(xq) = f(xk) • k nearest neighbor • Given query xq, estimate using majority (mean) of k nearest neighbors Intelligent Environments
Nearest Neighbor Intelligent Environments
Nearest Neighbor • Pros • Fast training • Complex target functions • No loss of information • Cons • Slow at query time • Easily fooled by irrelevant attributes Intelligent Environments
Bayes Classifier • Recall Bob example • D = training data • h = sample rule Intelligent Environments
Naive Bayes Classifier y represents Bob’s location • Naive Bayes assumption • Naive Bayes classifier Intelligent Environments
Bayes Classifier • Pros • Optimal • Discrete or numeric attribute values • Naive Bayes easy to compute • Cons • Bayes classifier computationally intractable • Naive Bayes assumption usually violated Intelligent Environments
Decision Tree Induction Day Sun M…F Sat Time > 0600 yes no Time < 0700 yes no Locationt … Bedroom Bathroom Intelligent Environments
Decision Tree Induction • Algorithm (main loop) • A = best attribute for next node • Assign A as attribute for node • For each value of A, create descendant node • Sort training examples to descendants • If training examples perfectly classified, then Stop, else iterate over descendants Intelligent Environments
Decision Tree Induction • Best attribute • Based on information-theoretic concept of entropy • Choose attribute reducing entropy (~uncertainty) from parent to descendant nodes A1 A2 v1 v2 v1 v2 Bathroom (25) Kitchen (25) Bathroom (25) Kitchen (25) Bathroom (50) Kitchen (0) Bathroom (0) Kitchen (50) ? ? B K Intelligent Environments
Decision Tree Induction • Pros • Understandable rules • Fast learning and prediction • Cons • Replication problem • Limited rule representation Intelligent Environments
Other Prediction Methods • Hidden Markov models • Radial basis functions • Support vector machines • Genetic algorithms • Relational learning Intelligent Environments
Prediction Issues • Representation of data and patterns • Relevance of data • Sensor fusion • Amount of data Intelligent Environments
Prediction Issues • Evaluation • Accuracy • False positives vs. false negatives • Concept drift • Time-series prediction • Distributed learning Intelligent Environments