|
|
|
|
| LEADER |
04405nam a2200229uu 4500 |
| 005 |
20190627124404.0 |
| 008 |
s2005 ci a |||||||||| ||eng|d |
| 035 |
|
|
|a HR-ZaFER 33928
|
| 040 |
|
|
|a HR-ZaFER
|b hrv
|c HR-ZaFER
|e ppiak
|
| 041 |
|
|
|a eng
|
| 080 |
|
|
|a 621.39
|h ELEKTROTEHNIKA
|j TELEKOMUNIKACIJE. TELEUPRAVLJANJE (DALJINSKO)
|e 621.3
|9 1248
|
| 100 |
1 |
|
|9 18140
|a Ferišak, Marko
|
| 245 |
|
|
|a Intelligent automata models in data mining processes for telecommunications service forecasting = Modeli inteligentnih automata u procesima dubinskih analiza predviđanja telekomunikacijskih usluga :
|b doctoral thesis = doktorska disertacija /
|c Marko Ferišak ; [mentor Mladen Tkalić]
|
| 260 |
|
|
|a Zagreb :
|b M. Ferišak ; Fakultet elektrotehnike i računarstva,
|c 2005.
|
| 300 |
|
|
|a 235 str. :
|b ilustr. ;
|c 30 cm +
|e CD
|
| 504 |
|
|
|a Bibliografija str. 218-229.
|
| 520 |
|
|
|a Skladištenje podataka dobiva sve važniju ulogu u telekomunikacijama. Podrška procesu poslovnog odlucivanja putem stvaranja izvještaja i analize omogucuje managementu donošenje odluka temeljenih na pravim podacima iz poslovne sredine u kojoj rade. Dubinske se analize mogu primjeniti pomocu raznih alata baziranih na inteligentnim automatima ili algoritmima. Ova disertacija predstavlja potpuni, od kraja do kraja, pristup dubinskim analizama, usmjeren na telekomunikacijsku okolinu. U disertaciji se prvo prezentiraju razni algoritmi, metode usporedbe modela, detalji važni za telekomunikacije, onda se prelazi na poslovnu pozadinu, implementaciju procesa dubinskih analiza u radnoj stanici koja se sastoji od baze podataka, interface-a, vlastitih neuronskih mreža i alata Weka. Radna stanica razvijena u okviru ove disertacije predstavlja jedan pristup procesu dubinske analize u telekomunikacijskoj marketinškoj okolini s ciljevima smanjenja troškova i poboljšanog razumijevanja modela poslovnog procesa. Ova je radna stanica testirana na više testnih skupova podataka s ciljem usporedbe nekoliko algoritama, uzevši u obzir njihovu snagu u procesima dubinske analize i prakticne aspekte njihove primjene.
Ključne riječi: inteligentni automati, skladištenje podataka, dubinske analize, Weka, neuronske mreže, neuronski automati, telekomunikacije, radna stanica za dubinske analize
|
| 520 |
|
|
|a Data warehousing role in telecommunications continues to increase in importance. Supporting business decision making through reports and analysis Data Warehouse empowers management to act based on real numbers from the business environment they operate in. Data mining processes play an important role in analytical processes, either for the purpose of decision making support or marketing campaign support.
In the field of marketing, especially given the environment full of various information and offers, personalized approach becomes a must. Here data warehousing and data mining offers the technology required to implement personalized marketing approach, allowing for the better understanding of the customer’s needs and preferences.
Data mining can be implemented through different tools offering intelligent automata or various algorithms. This thesis presents a complete end to end approach to data mining aimed at a telecommunications environment. It starts with the presentation of various algorithms, ways to compare models, details important from the telecommunication standpoint, then moves on to the business background, mining process implementation in a workstation consisting of database, interfaces, self developed neural network and open source tool Weka.
Workstation delivered in the scope of this thesis presents an approach to data mining in telecommunications marketing campaign environment with the goals of cost down and improved business model understanding. This workstation is then tested on multiple test data sets in order to compare several algorithms both from their mining capability and from the practical aspects of their implementation. Final conclusion from the tests is that no single algorithm is the best for all situations, but an analyst is required to interact with the process, at least occasionally in order to perform verification.
Key Words: intelligent automata, data warehousing, data mining, Weka, neural networks, neural automata, telecommunications, data mining workstation
|
| 700 |
|
|
|4 ths
|9 5453
|a Tkalić, Mladen
|
| 942 |
|
|
|c D
|2 udc
|
| 990 |
|
|
|a 31731
|
| 999 |
|
|
|c 29532
|d 29532
|