Python
TensorFlow, Keras, Scikit-learn, pandas
A two-layer Neural Netowrk where first layer is of 100 neurons, 2nd layer has 64 neurons and the last layer isthe prediction layer. We used Rectified Linear Unit for first two layers and Sigmoid function for the last layer for prediction.
A student dataset collected from CSE 35th Batch, AUST which is given in "Extended_1.csv".
-
Firstly we people's coffee drinking habit with their respective coding hours. Here, we found out that people who drink more coffee tend to code more. It is justifiable as we know that coders tend to drink more coffee.
-
We checked people's sleeping habit with their coffee drinking habit. It is noticed that people who sleep less than
$6$ hours tend to drink more coffee than people who sleep more which is also justifiable as coffee has more caffeine which prevents sleepiness. -
Next, we compared weight and coffee drinking habit. Basically we compared BMI(Body Mass Index) with coffee drinking habit. It shows that, people who drink more coffee tend to have higher BMI.
-
Later, we compared coffee drinking habit based on males and females. We found out that more males are addicted to coffee than females. 25% male drink coffee whereas 37% female don't drink coffee.
-
When compared coffee drinking with healthy food eating habit, it is seen that more regular food eaters drink coffee.
-
Lastly, we checked people's coffee drinking habits with their hometown. We categorized everyone's hometown based on division. We found out that, people whose hometown are in Barisal tend to drink the highest amount of coffee. On the other hand, people who are from Rangpur tend to drink less coffee.