Vehicle Classification Research Paper

Abstract

The vehicle classification, which consists of determining the vehicle of different company, is very important for a customer because a vehicle may fit into multiple categories. Before buying a vehicle we consult reviews, ratings from numerous agencies.

Some agencies perform rigorous testing on the vehicles and quantize the vehicle features like acceleration, braking, fuel economy etc, while other relies on the consumer reviews, awards won by particular vehicle model. Therefore mathematically we can say that: Vehicle success = ƒ (Vehicle features). In this work, we propose a new approach for vehicle classification based on a Probabilistic Neural Network and feature selection. Our goal is to classify a customer liking vehicle among

Accordingly, a PNN learns more quickly than many neural networks model and have had success on a variety of applications. Based on these facts and advantages, PNN can be viewed as a supervised neural network that is capable of using it in system classification and pattern recognition. The main objective of this paper is to describe the use of PNN in vehicle classification.

The architecture of PNN:

The PNN was first proposed in [11]. A probabilistic neural network is built with four layers as shown in the figure.1.

PNN is one of the types of neural networks with a one pass learning algorithm. . The PNN architecture is composed of many interconnected processing units or neurons organized in successive layers. The input layer unit does not perform any computation and simply distributes the input to the neurons in the pattern layer. On receiving a pattern from the input layer, the neuron of the pattern layer computes its output as the probability density function (pdf) for a single sample. i.e.,  

 

2

2

2

2 2.

1 

 

X

1 

 

Where ni is the total number of samples in the ith population.

If the a priori probabilities for each class are the same, and the losses associated with making an incorrect decision for each class are the same, the decision layer unit classifies the pattern in accordance with the Bayes’s decision rule based on the output of all the summation layer neurons.

i.e, hx g x i m i  argmax , 1,2,3,, Where hx denotes the estimated class of the pattern x and m is the total number of classes in the training samples. This includes determining the network size, the pattern layer neurons and an appropriate smoothing parameter.

(Fig-1)

International Journal of Application or Innovation in Engineering & Management (IJAIEM)

Web Site: www.ijaiem.org Email: [e-mail], [e-mail]

Volume 2, Issue 7, July 2013 ISSN 2319 - 4847

Volume 2, Issue 7, July 2013 Page 369

Training Set

The training set must be thoroughly representative of the actual population for effective classification:

 More demanding than most NN’s

 Sparse set sufficient

 Erroneous samples and outliers tolerable

Adding and removing training samples simply involves adding or removing “neurons” in the pattern layer