Browsing by Author "Rizvee, Md. Arif"

Now showing 1 - 3 of 3

Content Based Document Classification Using Soft Cosine Measure
(International Journal of Advanced Computer Science and Applications, 2019) Hasan, Md Zahid; Hossain, Shakhawat; Rizvee, Md. Arif; Rana, Md. Shohel
Abstract: Document classification is a deep-rooted issue in information retrieval and assumed to be an imperative part of an assortment of applications for effective management of text documents and substantial volumes of unstructured data. Automatic document classification can be defined as a content-based arrangement of documents to some predefined categories which is for sure, less demanding for fetching the relevant data at the right time as well as filtering and steering documents directly to users. For recovering data effortlessly at the minimum time, scientists around the globe are trying to make content-based classifiers and as a consequence, an assortment of classification frameworks has been developed. Unfortunately, because of using conventional algorithms, almost all of these frameworks fail to classify documents into the proper categories. However, this paper proposes the Soft Cosine Measure as a document classification method for classifying text documents based on its contents. This classification method considers the similarity of the features of the texts rather than making their physical compatibility. For example, the traditional systems consider ‘emperor’ and ‘king’ as two different words where the proposed method extracts the same meaning for both of these words. For feature extraction capability and content-based similarity measure technique, the proposed system scores the classification accuracy up to 98.60%, better than any other existing systems.
Machine Learning approach for predicting temperature in Bangladesh
(Daffodil International University, 2018-12) Rizvee, Md. Arif
Our country has six seasons. Though temperature remain high during summer season in Bangladesh. For global worming Bangladesh weather is also changing. As a result temperature in Bangladesh is increasing day by day. For this reason there are many natural disasters like cyclone, flood, drought are happening in Bangladesh regularly. As a result so many problems like death of people, food shortage, adequate of pure drinking water are arising. But currently Bangladesh doesn’t have any modern temperature prediction technique. It is difficult to predict the weather temperature due to non-linear characteristics of natural calamities in the country. It is proved that artificial intelligence is very helpful for predicting weather temperature. We use machine learning technique for predicting temperature of weather. Linear regression technique is a popular method for predicting temperature. Besides temperature, humidity, rain fall, dew point is used as predictors.
Weather Forecasting for the North-Western region of Bangladesh
(Scopus, 2020) Rizvee, Md. Arif; Arju, shfakur Rahman; Hasan, Md. Al-; Tareque, Saifuddin Mohammad; Hasan, Md.Zahid
Weather forecasting has several effects in our everyday life from farming to event planning. In the northwestern part of Bangladesh, various natural calamities cause the tragic death of many people and economic loss which impacts the total economic growth in Bangladesh. However, the nonlinear relationship between the input parameters and output data in the weather forecasting system makes it more complex. This study investigates the machine learning-based weather forecasting model for the north-western part of Bangladesh to enhance the accuracy of forecasting results in short periods. Artificial neural networks and extreme learning machine algorithms were used for a strong weather prediction purpose. In this experiment, thirty years of historical weather data of temperature, rain, wind, and humidity from seven weather stations in the northwestern part were collected from the Bangladesh Meteorological Department (BMD). The Extreme Machine Learning (ELM) model performs better than Artificial Neural Network and the accuracy rate is 95%.