(21th-September-2020) • DNN is hard to learn - over learning • It is okay if you do "pre-training per layer"! [Hinton + 06] •...

(20th-September-2020) History of Neural network is following time line. Difficulty in learning • Over learning: The learning error is...

(19th-September-2020) Classify quiz is here. Then, each function are following.

(18th-September-2020) Other case study are following. • Speech recognition • Issues: Input: MFCC from voice Output: HMM state past...

(17th-September-2020) Deep learning is a subset of machine learning in artificial intelligence (AI) that has networks capable of learning...

(16th-September-2020) • Probability can be used to make decisions under uncertainty. • The posterior probability is used to update an...

(15th-September-2020) • A dynamic belief network (DBN) is a belief network with regular repeated structure. It is like a (hidden) Markov...

(14th-September-2020)

(13th-September-2020)

(12th-September-2020) • A hidden Markov model (HMM) is an augmentation of the Markov chain to include observations. Just like the state...

(11th-September-2020) • model a dynamic system as a belief network by treating a feature at a particular time as a random variable. We...

(10th-September-2020) • For example, using particle filtering to compute P(Report|smoke) for the belief network of Figure 6.1. First...

(9th-September-2020) Instead of creating a sample and then rejecting it, it is possible to mix sampling with inference to reason about...

(8th-September-2020) Rejection function Given some evidence e, rejection sampling estimates P(h|e) using the formula P(h|e) = (P(h...

(7th-September-2020) • Probabilities can be estimated from a set of examples using the sample average. The sample average of a...

(6th-September-2020) • To generate samples from a single discrete or real-valued variable, X, first totally order the values in the...

(5th-September-2020) • This section gives an algorithm for finding the posterior distribution for a variable in an arbitrarily structured...

(4th-September-2020) • The most common probabilistic inference task is to compute the posterior distribution of a query variable given...

(3rd-September-2020) • This example also illustrates another example of explaining away and the preference for simpler diagnoses over...

(2nd-September-2020) • To represent a domain in a belief network, the designer of a network must consider the following questions: • What...