We have a group. The group has made the conclusion of the project. I will send two screenshots of the conclusion and the title of the project, because the teacher said to give a presentation tonight. Each member of the group will give a speech, and I am assigned to the speech conclusion part. You can prepare a speech for me for about 5 minutes through this conclusion part. Thank you experts, Probably all the speeches were sent to me in written form. thank you!

The impacts of a well-balanced diet on immunity in combating the COVID-19 virus in various countries Questions: 1. How many countries adhere to the health authorities' recommendations to consume at least 40% vegetables for a balanced diet? 2. How is the current death and total cases in countries that consume recommended amount of vegetables for their daily intakes as compare to those that doesn't follow the recommendation? 3. Compare different machine learning models to identify the predictive COVID-19 trends in countries where sufficient vegetables is consumed or vice versa? Conclusion: For our third question we wanted to compare the different machine learning models to identify the best model to use for prediction in similar use cases to predict the timesearies. In our case we had use Randomforest and Linearregression. We had used the Top 3 and Bottom 3 countries according to the vegetable consumption. Our results as shown in following table clearly states that Randomforest has better accuracy and performance as compare to the linearregression. As an exmaple we got the 46.27% accuracy of predicting covid deaths in "Madagascar" using Linearregression while Randomforest has outperfomed and gives mean absolute error of 0.7 which means 99.3% accuracy. Though the numbers from Randomforest are verry efficient yet in some cases Randomforest does underperformed as in the case of India where the mean absolute value is abut 34.3 which is about 76.7% accuracy not as high as other countries. It might further be improved by fixing the model paramters and find the best suitbale paramters that can perform well in every case for this use case. Accuracy (Linear regression) Mean Absolute Error (Random Forest) Country India 90.21 30.31 Tajikistan 15.832 0.41 Armenia 66.497 2.21 Tunisia 66.79 6.46 Haiti 21.68 0.93 Panama 60.58 3.25 Madagascar 46.27 0.70 Some other imporatnt factors that cause the models to result in low performance or accuracy is the source of data and some countries doesn't release the accurate data as oppose to others. These factors give highly diversified results, it can also be improved with higher accuracy of data. Due to limitation of time and resources, we able to scrab the covid data for shoarter period of time. We can further improve this by having better resource and larger amount of dataset