Simple Naïve Bayes Model
This model tries to predict smartphone type (iPhone vs. Android) based on several characteristics and personality traits like gender, age, honesty/humility, extraversion, conscientiousness, avoidance similarity, emotionality, agreeableness, socio-economic status, time owned current device, and whether the phone is perceived as a status object. The accuracy of this model is 66.7.3%, meaning that the model can correctly predict the type of smartphone (Android / iPhone) about 67% of the time. With the mosaic plot you can see the device preferences according to gender.
Overall accuracy = 0.667
Confusion matrix
Predicted (cv)
Actual Android iPhone
Android 0.543 0.457
iPhone 0.245 0.755
Naïve Bayes Model with Kernels
The NB model assumes a normal/gaussian distribution, but Naive Bayes will it will also allow you to estimate the mean of each feature empirically using kernels. This will estimate quantiles of the distribution, which might help when you have skewed distributions; on the other hand, lots of observations are needed in order to get reliable estimates. In this case, 200-300 observations in each group might not be enough to do well. With this model, accuracy raises from 67% to roughly 70%. Moreover, you have a 56% chance of getting Android users right, and 80% chance of getting iPhone users right. In the plot you can see the device preferences according to age, with a spike for younger adults choosing the iPhone over an Android device.
Overall accuracy = 0.699
Confusion matrix
Predicted (cv)
Actual Android iPhone
Android 0.557 0.443
iPhone 0.200 0.800