Upload presentasi
Presentasi sedang didownload. Silahkan tunggu
Diterbitkan olehYanto Azis Telah diubah "10 tahun yang lalu
1
Collabnet Overview v 1.2 021201 Informatika Introduction
2
Informatika What is an ANN? It is a computational System Inspired by Structure, Processing method, Learning ability of Biological Brain The term “neural network” is also applied to models of the brain
3
Informatika Why ANN? Some tasks can be done easily (effortlessly) by humans but are hard by conventional paradigms on Von Neumann machine with algorithmic approach Computers can perform many operations considerably faster than a human being
4
Informatika Why ANN? Massive Parallelism Distributed representation Learning ability Generalization ablity Fault tolerance
5
Informatika Characteristics of ANN A large number of very simple processing neuron-like processing elements A large number of weighted connections between the elements Distributed representation of knowledge over the connections Knowledge is acquired by network through a learning process
6
Informatika Biological Neuron Dendrit, bertugas menerima informasi Soma, tempat pengolahan informasi Axon, mengirim inpuls-inpuls ke sel syaraf lainya Synapse, penghubung antara 2 neuron
7
Informatika Biological vs Artificial Otak ManusiaJST SomaNode DendritesInput/Masukan AxonOutput/Keluaran SynapsisWeight/ Bobot
8
Informatika Artificial Neuron Σ p2p2...... SUM w1w1 w2w2 wiwi Weight F(y) n=Σp i.w i a=f(n) Activation Function
9
Informatika Neuron vs Node
10
Informatika Topology
11
Informatika Learning Learn the connection weights from a set of training examples Different network architectures required different learning algorithms
12
Informatika Supervised Learing The network is provided with a correct answer (output) for every input pattern Weights are determined to allow the network to produce answers as close as possible to the known correct answers The back-propagation algorithm belongs into this category
13
Informatika Unsupervised Learning Does not require a correct answer associated with each input pattern in the training set Explores the underlying structure in the data, or correlations between patterns in the data, and organizes patterns into categories from these correlations The Kohonen algorithm belongs into this category
14
Informatika Applications Pattern Classification Clustering/Categorization Function approximation Prediction/Forecasting Optimization Content-addressable Memory Control
15
Informatika Sejarah Model JST formal pertama diperkenalkan oleh McCulloch dan Pitts (1943) 1949, Hebb mengusulkan jaringan Hebb 1958, Rosenblatt mengembangkan perceptron untuk klasifikasi pola 1960, Widrow dan Hoff mengembangkan ADALINE dengan aturan pembelajaran Least Mean Square (LMS) 1974, Werbos memperkenalkan algoritma backpropagation untuk perceptron banyak lapisan
16
Informatika Sejarah 1975, Kunihiko Fukushima mengembangkan JST khusus pengenalan karakter, disebut cognitron, namun gagal mengenali posisi atau rotasi karakter yang terdistorsi 1982, Kohonen mengembangkan learning unsupervised untuk pemetaan 1982, Grossberg dan Carpenter mengembangkan Adaptive Resonance Theory (ART, ART2, ART3) 1982, Hopfield mengembangkan jaringan Hopfield untuk optimasi
17
Informatika Sejarah 1983, perbaikan cognitron (1975) dengan neocognitron 1985, Algoritma Boltzmann untuk jaringan syaraf probabilistik 1987, dikembangkan BAM (Bidirectional Associative Memory) 1988, dikembangkan Radial Basis Function
18
Collabnet Overview v 1.2 021201 Informatika Thank’s Any Questions ?
Presentasi serupa
© 2024 SlidePlayer.info Inc.
All rights reserved.