Scribd
Upload
109 views3 pages

Teme Pentru Referate La Cursul "Retele Neuronale"

This document provides 15 potential topics for course papers on neural networks and genetic algorithms. Some of the proposed neural network topics include perceptrons and their applications in digit recognition, feedforward neural networks and their use in handwritten text and image recognition. Genetic algorithm topics involve applying genetic algorithms to optimization problems like the traveling salesman problem and knapsack problem. The document provides keywords for each topic and guidelines for writing the course papers, which should be 5-8 pages and cite at least 3 relevant sources. Students will be evaluated based on their course paper (60% of grade) and in-class work (40% of grade).

Uploaded by

cernatand
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
109 views3 pages

Teme Pentru Referate La Cursul "Retele Neuronale"

This document provides 15 potential topics for course papers on neural networks and genetic algorithms. Some of the proposed neural network topics include perceptrons and their applications in digit recognition, feedforward neural networks and their use in handwritten text and image recognition. Genetic algorithm topics involve applying genetic algorithms to optimization problems like the traveling salesman problem and knapsack problem. The document provides keywords for each topic and guidelines for writing the course papers, which should be 5-8 pages and cite at least 3 relevant sources. Students will be evaluated based on their course paper (60% of grade) and in-class work (40% of grade).

Uploaded by

cernatand
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
You are on page 1/ 3

Teme pentru referate la cursul Retele Neuronale

1. Perceptronul. Istoric. Aplicatii in clasificare (exemplu: recunoasterea cifrelor)


Keywords: perceptron, Rosenblatt, digit recognition
2. Sisteme de recunoastere a caracterelor. Structura generala. Exemple. Detalierea modulelor
bazate pe retele neuronale.
Keywords: optical character recognition, neural networks
3. Retele liniare. Arhitectura. Algoritmul Widrow-Hoff. Aplicatii in filtrare adaptiva.
Keywords: linear networks, Widrow-Hoff algorithm, adaptive filters, ADALINE
4. Retele feed-forward multi-nivel. Aplicatii in recunoasterea scrisului de mana.
Keywords: feedforward neural networks, handwritten text recognition
5. Retele feed-forward multi-nivel. Aplicatii in clasificarea imaginilor.
Keywords: feedforward neural networks, face recognition, texture classification
6. Variante ale algoritmului BackPropagation: varianta cu moment, RProp, QuickProp, variante
bazate pe metoda lui Newton
Keywords: Backpropagation algorithm, momentum variant, RProp, QuickProp, Netwton method
7. Retele RBF si retele probabiliste. Aplicatii in rezolvarea problemelor de clasificare
Keywords: radial basis function networks, probabilistic network, classification
8. Retele neuronale de tip ART. Aplicatii in gruparea datelor
Keywords: adaptive resonance theory, ART1, ART2, data clustering
9. Retele de tip Kohonen. Aplicatii in controlul bratelor de robot
Keywords: Kohonen networks, self-organizing maps, robot arm control
10. Retele de tip Kohonen. Aplicatii in categorizarea documentelor electronice
Keywords: Kohonen networks, self-organizing maps, WebSOM
11. Retele neuronale pentru discretizare vectoriala. Aplicatii in compresia imaginilor
Keywords: vector quantization, image compression
12. Retele neuronale pentru analiza in componente principale. Aplicatii in analiza imaginilor
Keywords: principal component analysis, Oja algorithm, Sanger algorithm, image processing
13. Aplicatii ale retelelor neuronale in detectia intrusilor in retele de calculatoare
Keywords: neural networks, intrusion detection
14. Aplicatii ale retelelor neuronale in analiza si predictie financiara
Keywords: neural networks, financial analysis, stock prediction
15. Aplicatii ale retelelor neuronale in proiectarea sistemelor de asistare a deciziei medicale
Keywords: neural networks, medical diagnosis

Teme pentru referate la cursul Algoritmi genetici


1. Aplicatii ale algoritmilor genetici in rezolvarea problemelor de optimizare combinatoriala.
Problema comis voiajorului
Keywords: genetic algorithms, traveling salesman problem
2. Aplicatii ale algoritmilor genetici in rezolvarea problemelor de optimizare combinatoriala.
Problema rucsacului
Keywords: genetic algorithms, knapsack problem
3. Aplicatii ale algoritmilor genetici in rezolvarea problemelor de optimizare combinatoriala.
Problema impachetarii
Keywords: genetic algorithms, bin packing problem
4. Aplicatii ale algoritmilor genetici in rezolvarea problemelor de planificare. Generarea automata
a orarelor
Keywords: genetic algorithms, timetabling
5-6. Aplicatii ale algoritmilor genetici in simularea proceselor creative (generarea automata de
"pictura"/"muzica")
Keywords: genetic algorithms, evolutionary art, computer generated music
7. Strategii evolutive cu auto-adaptare
Keywords: evolution strategies, self-adapting
8. Programare genetica. Aplicatii in sinteza automata a circuitelor
Keywords: genetic programming, circuits synthesis
9. Programare genetica. Aplicatii in proiectarea filtrelor pentru prelucrarea imaginilor
Keywords: genetic programming,image processing
10. Programare genetica. Aplicatii in analiza financiara
Keywords: genetic programming, financial analysis
11.Programare genetica. Aplicatii in generarea strategiilor de joc
Keywords: genetic programming, games strategies
12. Proiectarea evolutiva a retelelor neuronale
Keywords: evolutionary programming, neural networks
13-15. Inventar al bibliotecilor software open-source destinate proiectarii: (a) algoritmilor
genetici; (b) strategiilor evolutive; (c) programare genetica
Keywords: software framework/library/tool, genetic algorithms/evolution strategies/genetic
programming

Indicatii pentru redactarea referatului:


1. Referatul are caracter bibliografic si va reprezenta 60% din nota finala (restul de 40% va fi
reprezentat de lucrarea scrisa si activitatea pe parcursul semestrului). Nu este necesara
implementarea algoritmilor, insa realizarea unor implementari originale aduce bonus de pana la 4
puncte.
2. Referatul se va baza, pe langa materialele de la curs, pe cel putin trei materiale bibliografice
selectate folosind cuvintele cheie specificate la fiecare tema. Materialele bibliografice vor fi
folosite pentru a realiza o scurta sinteza. Bibliografia utilizata va fi obligatoriu mentionata la
sfarsitul referatului.
3. Nu se impune un anumit numar de pagini insa temele propuse ar trebui tratate in circa 5-8
pagini.
4. Lucrarea scrisa consta intr-un test grila cu intrebari din tematica cursurilor. Se pot folosi
materialele bibliografice (http://www.info.uvt.ro/~dzaharie/arad)
D. Zaharie

You might also like