Graded quiz of How to Win a Data Science Competition

ByAdmin July 11, 2021

Week – 4 Graded quiz How to Win a Data Science Competition Learn from Top Kagglers

n_estimators, max_depth, min_samples_split.

n_jobs, random_state, verbose.

bootstrap, oob_score, warm_start.

2. Suppose you fit LightGBM to your train data and check performance on the validation set. The train set consists of 500 rows and 1000 different features and validation set consist of 50 objects. You run automatic hyperparameter optimization method overnight and in the morning you select the best parameters, produce results for the test set and submit to the leaderboard. We also know that test set comes from the same distribution as train and validation sets.

1 point

There is a high chance of overfitting to the validation set. That is, there is a high chance that score on the test set will be bad. This is because we’ve tried too much hyperparameters while the dataset is small and the number of features is large.

There is a low chance of overfitting to the validation set. That is, there is a high chance that score on the test set will be good. This is because we found good parameters on validation set and test set is similar to the test set.

Leave-one-out validation (i.e. fit model for all points except one and estimate quality for this single point; repeat for every point).

Select model by quality on training set (i.e. fit model on whole dataset and measure quality on the same data).

k-Fold scheme (i.e. split data into k part, use k-1 parts for training and the last one for quality estimation; repeat for each part)

Hold-Out scheme (i.e. divide data into two parts, use first for model fitting and second for estimation of quality).

Add (or increase) Weight Decay.

Change optimization method to Adam.

Insert (or increase rate of) Dropout layers inside NN.

Add more layers.

Reduce number of parameters (e.g. remove some layers)

How to Win a Data Science Competition

Feature preprocessing and generation with respect to models
ByAdmin July 10, 2021

Week – 1 Feature preprocessing and generation with respect to models 1. Suppose we have a feature with all the values between 0 and 1 except few outliers larger than…

Read More Feature preprocessing and generation with respect to models
How to Win a Data Science Competition

All Assignment of How to Win a Data Science Competition
ByAdmin July 10, 2021

All Assignment of How to Win a Data Science Competition Learn from Top Kagglers Week – 1 Assignment Programming Assignment: Pandas basics Click Here To Get Week – 2 Assignment…

Read More All Assignment of How to Win a Data Science Competition
How to Win a Data Science Competition

Graded Advanced Features II Quiz
ByAdmin July 11, 2021

Week – 4 Graded Advanced Features II Quiz >>> How to Win a Data Science Competition Learn from Top Kagglers 1. Imagine that we apply X = PCA(n_components=5).fit_transform(data) and data…

Read More Graded Advanced Features II Quiz
How to Win a Data Science Competition

Mean encodings
ByAdmin July 11, 2021July 11, 2021

Week – 3 Mean encodings >>> How to Win a Data Science Competition: Learn from Top Kagglers 1. What can be an indicator of usefulness of mean encodings? 1 point…

Read More Mean encodings
How to Win a Data Science Competition

Exploratory data analysis
ByAdmin July 10, 2021

Week – 2 Exploratory data analysis >>> How to Win a Data Science Competition Learn from Top Kagglers 1. Suppose we are given a data set with features XX, YY,…

Read More Exploratory data analysis
How to Win a Data Science Competition

Data leakages
ByAdmin July 11, 2021

Week – 2 Data leakages >>> How to Win a Data Science Competition Learn from Top Kagglers 1. Suppose that you have a credit scoring task, where you have to…

Read More Data leakages

Week – 4 Graded quiz How to Win a Data Science Competition Learn from Top Kagglers

Similar Posts