No abstract

No abstract

Forecasting in pandemics and disasters is one of the means that contribute to reducing the damage of this pandemic, and the Corona virus is reportedly the most dangerous pandemic that the entire world is suffering from. As a result, we aim to use a deep learning algorithm to predict confirmed and new cases of Covid-19 in our study. This paper identifies the most essential deep learning techniques. Long short-term memory (LSTM) and gated recurrent unit (GRU) were shown to forecast verified Covid-19 fatalities in Malaysia, Egypt, and the U.S. using time series data from 1 January 2021 to 14 May 2022. The first section of this study examines a comparison of prediction models, while the second section examines how prediction and performance analysis may be enhanced using mean absolute error (MAE), mean absolute error percentage (MAPE), and root mean squared error (RMSE) Metrics. On the basis of the regression curves of two two-layer models, the data were split into training sets of 80% and test sets of 20%. The conclusion is that the outputs of the training model and the original data greatly converged. The findings of the study indicated that, for predicting Covid-19 cases, the GRU model in the three nations is superior than the LSTM model.

This paper presents a research study in developing lesson plans for differentiated learning or adaptive learning using "lesson" on Moodle LMS platform. The lesson are to provide learning paths based on inquiring category from Laurillard Conversational Framework. The design of various learning paths provide teaching methods used in class such as videos, quizzes and reading tasks. This lesson when planned using flowcharts would guide the instructors to link the topics and assessment in any order that would promote students' self-paced and self-learning experience.

No abstract