Malaria remains a major infectious disease that causes a great deal of damage, mainly in African countries. The aim of this work is to assess the dynamics of malaria in the Adamawa Region of Cameroon using time series approaches. This will serve as baseline for further improvement. To this end, data malaria on cases occurring between 2018 and 2022, collected weekly in the form of time series are used. After collecting the data and built up the dataset, the data has been first statistically described. Second, time series analysis is carried out on the dataset. Finally, forecasting is performed using selected basic models, along with a comparison of prediction accuracy metrics. For forecast, a statistical, a machine learning, a deep learning models are used, in addition with a built-up tool. They are ARIMA, RFR, LSTM and Facebook Prophet. Evaluation metrics involved in the benchmark analysis are MAE, RMSE and R2. Random Forest Regressor (RFR) gives the best scores in training for the three metrics, while LSTM gives the best in testing likewise. These prediction models result and data used will serve as benchmarks for the development of more elaborate and precise models, aimed ultimately at producing factual elements for decision support in rolling back Malaria.