DATA CLASSIFICATION SYSTEM WITH FUZZY NEURAL BASED APPROACH
Knowledge Discovery in Database and Data Mining use techniques derived from machine learning, visualization and statistics to investigate real world data. Their aim is to discover patterns within the data which are new, statistically valid, interesting and understandable. In recent years, there...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Year Project |
| Language: | English |
| Published: |
Universiti Teknologi Petronas
2005
|
| Subjects: | |
| Online Access: | http://utpedia.utp.edu.my/7616/1/2005%20-%20DATA%20CLASSIFICATION%20SYSTEM%20WITH%20FUZZY%20NEURAL%20BASED%20APPROACH.pdf http://utpedia.utp.edu.my/7616/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Knowledge Discovery in Database and Data Mining use techniques derived from
machine learning, visualization and statistics to investigate real world data. Their aim is
to discover patterns within the data which are new, statistically valid, interesting and
understandable.
In recent years, there has been an explosion in computation and information technology.
With it have come vast amounts of data. Lying hidden in all this data is potentially useful
information that is rarely made explicit or taken advantage. New tools based both on
clever applications of established algorithms and on new methodologies, empower us to
do entirely new things. In this context, data mining has arisen as an important research
area that helps to reveal the hidden interesting information from the rawdatacollected.
The project demonstrates how data mining can address the need of business intelligence
in the process of decision making. An analysis on the field of data mining is done to show
how data mining can help in business such as marketing, credit card approval. The
project's objective is identifying the available data mining algorithms in data
classification and applying new data mining algorithm to perform classification tasks.
The proposed algorithm is a hybrid system which applied fuzzy logic and artificial neural
network, which applies fuzzy logic inference to generate a set of fuzzy weighted
production rules and applies artificial neural network to train the weights of fuzzy
weighted rules for better classification results.
Theresult of this system using the iris dataset and credit card approval dataset to evaluate
the proposed algorithm's accuracy, interpretability. The project has achieved the target
objectives; it can gain high accuracy for data classification task, generate rules which can
help to interpret the output results, reduce the training processing. But the proposed
algorithm still require high computation, the processing time will be long if the dataset is
huge. However the proposed algorithm offers a promising approach to building
intelligent systems. |
|---|