Font size
WorksheetsModel Tuning Quiz
Total questions: 18
Worksheet time: 10mins
A model can be seen as a combination of :
Hyperparameters
Algorithm
Training Data
Testing Data
To validate a model, you use the :
Correcting set
Running set
Testing set
Training set
Model Tuning is the process of maximizing a model's performance by changing its (a)
Model tuning is :
a try-and-go process
a trial-and-error process
a choose-and-buy process
a satisfy-and-test process
The Ideal model have :
Low Variance
High Variance
Low Bias
High Bias
A model that "might always do good prediction, but never great prediction" have :
High Variance and High Bias
High Variance and Low Bias
Low Variance and High Bias
Low Variance and Low Bias
A model that "might do great sometimes, but it also might do terrible" have :
High Variance and High Bias
High Variance and Low Bias
Low Variance and High Bias
Low Variance and Low Bias
Bias represents ________ on a model's results due to erroneous assumptions in the machine learning process.
a systematic error
an occasional error
Variance, is a type of error that occurs due to a model's _____ to fluctuations in the training set.
rigidity
sensitivity
High variance would cause an algorithm to model the noise in the training set. This is most commonly referred to as over (a)
The Goal of model tuning is to find a model :
That is at least 95% accurate on the training data set
That takes less than 30 minutes to train and validate
That makes you feel like you are a true machine learning expert
That is at the sweet spot between a simple working model and a very complex one
Write down one mean for model tuning :
(a)
To which means this definition fits :
Optimization method that allow us to test a range of hyperparameters and to compare performances in order to deduce the best setting.
Grid Search
Grid Validation
Cross Search
Cross Validation
To which means this definition fits :
Technique used to test the effectiveness of a machine learning models and get more information about our algorithm performance.
Grid Search
Grid Validation
Cross Search
Cross Validation
A collection of individual models that learn to predict a target by combining their strengths and avoiding the weaknesses of each is called :
A Collection
A Group
An Ensemble
A Fusion
Bagging :
Is a parallel ensemble
Is a sequential ensemble
Aim to decrease variance, not bias
Aim to decrease bias, not variance
Boosting :
Is a circular ensemble
Is a sequential ensemble
Aim to decrease variance, not bias
Aim to decrease bias, not variance
Hyperparameters are :
Automatically set and adjusted
Arbitrarily set and adjusted
