Research
Primary Education Status Analysis in Bangladesh Based On Neural Networks and Baysian Networks
In this research work we have concentrate to measure the primary education status in Bangladesh, a developing country of South Asia. It known that the literacy rate of South Asian country is very slow and it is not the different in Bangladesh. Here we measure the dropout rate of primary school kids at different classes at different sessions. We have collected the data from various primary schools from Chittagong region of Bangladesh. Here we use K –Nearest Neighbor (KNN) algorithm to classify the data from irrelevant data like secondary school and tertiary level data. After then we have applied Neural Network (NN) to train the data set for better result. Finally we have compared the result by calculating the result with Bayesian Network (BN). Here we found that if the dropout rate is small Neural Network is best to measure the result and NN generate more error when the dropout rate is large. On the contrary BN is better when the rate is large.
Uncertainty Analysis for Spatial Image Extractions in the context of Ontology and Fuzzy C-Means Algorithm
This paper emphasis on spatial feature extractions and selection techniques adopted in content based image retrieval that uses the visual content of a still image to search for similar images in large scale image databases, according to a user’s interest. The content based image retrieval problem is motivated by the need to search the exponentially increasing space of image databases efficiently and effectively. It is also possible to classify the remotely sensed image to represent the specific feature of the target images. In this research we first imposed the Fuzzy C-means algorithm to our sample image and observed its value. After getting the experimental result from Fuzzy C-means we have had designed Ontological Matching algorithm which aftereffect better than the previous one. We have had espy that our Ontological Matching algorithm is twenty (20%) percent better than Fuzzy C-means algorithm.
