For advertisers and publishers, being able to forecast an ad's click-through rate with accuracy is essential since it enables them to target the appropriate audiences and place ads where they will have the greatest impact.
Advertisers and publishers can increase the efficiency of their advertising campaigns and maximise their return on investment by estimating the likelihood that an ad will be clicked or not, and by doing so, can decide which ads to show to which viewers.
While exploring the dataset provided for the challenge, there were two major problems we faced, one was the percentage of null values in the dataset and other was the problem of Data Imbalance in the Predictive Column. The null values can be solved by computing the average, mean, median whereas the imbalance can be solved using Random Forest Algorithm and Stacking.
The dataset contains large null values. It can bias the results of the machine learning models or reduce the accuracy of the model. So, we have to encode various values of variables in the dataset to deal with these null values.
The main problem that comes in random forest is that a large number of trees can make the algorithm too slow and ineffective for real-time predictions. In general, these algorithms are fast to train, but quite slow to create predictions once they are trained.
The benefit of stacking is that it can harness the capabilities of a range of well-performing models on a classification or regression task and make predictions that have better performance than any single model in the ensemble
Humans aren't well equipped to compute the meaning of multiple values abstracted in visual form. When visualizations include too much data, information overwhelms, and data melts into a graphic soup that most viewers can't stomach.
Tracks Applied (1)
Discussion