A Data Driven Approach to Forecasting BangladeshNext Generation Economy

Image

The issues of GDP have gotten the most worried among macroeconomic factors and information on GDP is viewed as the significant file for evaluating the public financial turn of events and for making a decision about the working status of the macro economy in general. It is crucial to Forecast microeconomic variables in the economic terminology. The main macroeconomic factors to gauge are the Gross Domestic Product (GDP), swelling and joblessness. As a total proportion of absolute financial creation for a country, GDP is one of the essential markers used to gauge the nation's economy. Since significant monetary and political choices depend on conjectures of these macroeconomic factors, it is basic that they are just about as solid and exact as could be expected. Erroneous figures might bring about destabilizing strategies and a more unstable business cycle. GDP is quite possibly the main pointers of public financial exercises for a nation

A Machine Learning Approach for Sentiment Analysis of Customer Satisfaction of Bangladeshi Delivery Services

Image

Bangladesh has a large population which is causing the delivery system growing up day by day. Therefore, some companies who provide these delivery services usually called “Currier Service” are growing gradually. In this paper, we choose the top delivery service company of Bangladesh and analyzed customers’ sentiment based on reviews and comments which is collected from their social media pages. We collected data of customers comments from the “Sundorban Currier Services”, “Redx Currier Services” and “Pathao Delivery System” official social media page are verified posts. In this sentiment review, we have used Natural Language Processing (NLP) in Bangla to categorize those reviews using multiple machine learning models. In those models, we have tried to discover the number of negative and positive reviews or comments in real- life social media platforms and predicted the reviews that are negative or positive using the Unigram, Bigram, and Trigram methods.We have found that the Bigram feature is the best for this analysis because it has the highest accuracy of 90.72% and according to that the highest F1 score is 91.26. Using the Bangla NLP approach, the machine learning models are categorized in negative and positive reviews with the sentiment analysis method.

Cryptocurrency Market Research using Machine Learning Approach

Image

For preventing counterfeiting and double-spending the encrypted digital or virtual money is used, which is called Cryptocurrencies. They are applied online exclusively. Cryptocurrency like bitcoin used peer to peer connection. In real world, these cryptocurrencies have no physical existence. They have no visible presence. There has no authority of the government over cryptocurrency. Functioning cryptocurrency relies on a technology called a blockchain. This blockchain was founded to relieve the problem of double-spending and also the interrupt in the centralized parties; control in the assets transaction. It is Bitcoins most significant invention. For keeping track of all economic and financial transactions the blockchain is used. This blockchain uses a cluster of computers. It can be said simply that, this technology is so strong that it can keep records permanently of transactions of business, assets, financial data, contract conversion, and property which is intellectual.