Introducing
Prezi AI.
Your new presentation assistant.
Refine, enhance, and tailor your content, source relevant images, and edit visuals quicker than ever before.
Trending searches
Loading…
Transcript
Classification between clean water and polluted water based on the electrical capacitance measurement using SCG Artificial Neural Network
NOR LIYANA BINTI JAMBARI
2017668752
SUPERVISOR : Ts Dr Mohd Suhaimi Bin Sulaiman
INTRODUCTION
- This paper demonstrates the application of Scaled Conjugate Gradient (SCG) of Artificial Neural Network (ANN) to classify between the clean water and polluted water using the cylindrical capacitance transducer measurement.
INTRO
Overview
Objectives
Problem Statement
Problem Statement
- The problem of water pollution is now becoming more serious.
- Polluted water can contribute to a serious risk of health and inhabitants.
- There are several conventional methods have involve in analysing the water sample.
- However, the method are found to be expensive and hard-handling
METHOD
PRODUCT #1
PRODUCT #1
PRODUCT #2
PRODUCT #2
PRODUCT #3
PRODUCT #3
RESULT & DISCUSSION
Cylindrical Capacitance Transducer Design.
Statistical Analysis
Classification using ANN
Model Neuron3 was selected to be the most optimized model for a single input capacitance measurement.
Meanwhile, the least optimized model is Neuron43
ANN
Conclusion
- By statistical analysis, the result shown a very distinguishable value for both cases.
- The result concluded that the data is normally distributed using the normality test by analyzing the Kolmogorov-Smirnov algorithm when the p-value is greater than 0.05.
- The Histogram and the Normal Q-Q plot also shows the claim that the research are normally distributed by analyze at the bell-shaped of the Histogram and the Q-Q plot on the linear line in the chart
- There was no overlap in the error bar
- For the ANN classification, the most optimized model selected for each input category can be shown to be significant
- The test confusion matrix showed that the system was accurate to 100% of sensitivity, specificity and accuracy of model Neuron3. Then, the result of all statistical analysis are analysis and the performance of confusion matrix.
Choose a template
Presentations from around the world
Learn more about creating dynamic, engaging presentations with Prezi