01.11.2014 Views

MACHINE LEARNING TECHNIQUES - LASA

MACHINE LEARNING TECHNIQUES - LASA

MACHINE LEARNING TECHNIQUES - LASA

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

4<br />

6.6.1 Weights bounds ..................................................................................................... 135<br />

6.6.2 Weight decay ......................................................................................................... 136<br />

6.6.3 Principal Components ............................................................................................ 137<br />

6.6.4 Oja’s one Neuron Model ........................................................................................ 139<br />

6.6.5 Convergence of the Weights Decay rule ............................................................... 139<br />

6.7 Anti-Hebbian learning ................................................................................................. 140<br />

6.7.1 Foldiak’s models .................................................................................................... 141<br />

6.7.2 CCA Revisited ....................................................................................................... 143<br />

6.7.3 ICA Revisited ......................................................................................................... 145<br />

6.8 The Self-Organizing Map (SOM) ................................................................................ 147<br />

6.8.1 Kohonen Network .................................................................................................. 147<br />

6.8.2 Bayesian self-organizing map ................................................................................ 150<br />

6.9 Static Hopfield Network .............................................................................................. 151<br />

6.9.1 Hopfield Network Structure .................................................................................... 152<br />

6.9.2 Learning Phase ...................................................................................................... 152<br />

6.9.3 Retrieval Phase ..................................................................................................... 153<br />

6.9.4 Capacity of the static Hopfield Network ................................................................. 154<br />

6.9.5 Convergence of the Static Hopfield Network ......................................................... 155<br />

6.10 Continuous Time Hopfield Network and Leaky-Integrator Neurons ...................... 155<br />

7. 7 Markov-Based Models ......................................................................................... 160<br />

7.1 Markov Process .......................................................................................................... 160<br />

7.2 Hidden Markov Models ............................................................................................... 161<br />

7.2.1 Formalism .............................................................................................................. 161<br />

7.2.2 Estimating a HMM ................................................................................................. 163<br />

7.2.3 Determining the number of states .......................................................................... 166<br />

7.2.4 Decoding an HMM ................................................................................................. 167<br />

7.2.5 Further Readings ................................................................................................... 169<br />

7.3 Reinforcement Learning ............................................................................................. 170<br />

7.3.1 Principle ................................................................................................................. 171<br />

7.3.2 Defining the Reward .............................................................................................. 172<br />

7.3.3 Markov World ......................................................................................................... 173<br />

7.3.4 Estimating the policy .............................................................................................. 174<br />

7.3.5 Summary ............................................................................................................... 175<br />

8. 8 Genetic Algorithms .............................................................................................. 177<br />

8.1 Principle ....................................................................................................................... 178<br />

8.2 Encoding ...................................................................................................................... 178<br />

8.3 Breeding and Selection .............................................................................................. 179<br />

8.4 The Algorithm .............................................................................................................. 181<br />

8.5 Convergence ............................................................................................................... 181<br />

9. 9 Annexes ................................................................................................................ 182<br />

9.1 Brief recall of basic transformations from linear algebra ....................................... 182<br />

9.1.1 Eigenvalue Decomposition .................................................................................... 182<br />

9.1.2 Singular Value Decomposition (SVD) .................................................................... 184<br />

9.1.3 Frobenius Norm ..................................................................................................... 184<br />

9.2 Recall of basic notions of statistics and probabilities ............................................ 184<br />

9.2.1 Probabilities ........................................................................................................... 184<br />

© A.G.Billard 2004 – Last Update March 2011

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!