Add estimated train time in Optuna mode

[pplonski]

In Optuna mode the train time is:

• Optuna tuning for each algorithm = len(algorithms)*otpuna_time_budget,
• ML model training = total_train_time

[pplonski]

For Optuna mode there is an additional print-out:

Expected computing time:
Total training time: Optuna + ML training = 3620 seconds
Total Optuna time: len(algorithms) * optuna_time_budget = 20 seconds
Total ML model training time: 3600 seconds