Your data shows a notable difference between the mean and the median values. Which type of scaling would be least affected by this discrepancy?
- All scaling methods are affected by this discrepancy
- Min-Max scaling because it scales all values between 0 and 1
- Robust scaling because it uses median and quartile ranges
- Z-score standardization because it creates a normal distribution
Robust scaling uses the median and interquartile range to scale the data, so it is not affected by the mean and is thus least affected by a discrepancy between the mean and the median.
In a scatter plot, the _____ and _____ of the dots represent the values of two different variables.
- color, size
- position, color
- position, size
- shape, color
In a scatter plot, the position of a dot on the X (horizontal) and Y (vertical) axis represents the values of two different variables. By looking at how the dots are scattered on the plot, one can deduce the type and strength of the relationship between the two variables.
In _________, the probability of an observation being missing is unrelated to both observed and unobserved data.
- All missing data
- MAR
- MCAR
- NMAR
In MCAR (Missing Completely at Random), the missingness is unrelated to both observed and unobserved data.
You are working in a clinical trial and your role is to confirm a certain hypothesis related to the drug effectiveness. Which type of data analysis should you focus on?
- All are equally suitable
- CDA
- EDA
- Predictive Modeling
CDA would be the most suitable as it focuses on confirming pre-formulated hypotheses, which in this case relates to the effectiveness of a drug.
The type of data that can be divided into categories but cannot be ordered or measured is _____.
- Nominal
- Ordinal
- Qualitative
- Quantitative
Nominal data can be divided into categories but these categories cannot be ordered or measured.
What does EDA stand for in the context of data analysis?
- Expanded Data Analysis
- Exploratory Data Analysis
- Exponential Data Analysis
- Extreme Data Analysis
EDA stands for Exploratory Data Analysis. This is a data analysis approach that involves the application of diverse techniques to gain insights about a dataset. Unlike classical methods, which usually begin with a preconceived hypothesis, EDA allows the data to speak for itself. It often involves summarizing the data, visualizing these summaries and looking for patterns, unusual observations, or inconsistencies that could inspire model building.
During the 'communicate' step of the EDA process, your audience is having difficulty understanding your conclusions. How could you address this issue?
- Adjust your communication approach to better meet the audience's understanding.
- Clarify their doubts during the communication phase.
- Ignore their difficulty and continue with the communication.
- Tell them to refer to the raw data for clarification.
If the audience is having difficulty understanding the conclusions during the 'communicate' phase, the best approach would be to adjust your communication to better meet the audience's understanding. This might involve simplifying complex concepts, using more visual aids, or providing more contextual explanations. Effective communication is key to ensuring the insights from the analysis are understood and can be acted upon.
What is the purpose of feature selection in machine learning?
- All of the above
- To identify and remove unimportant features
- To improve accuracy and speed of a machine learning model
- To reduce overfitting
The purpose of feature selection is to improve accuracy and speed of a machine learning model, reduce overfitting, and identify and remove unimportant features.
In the context of a Binomial Distribution, a "success" is defined as _____.
- a positive outcome
- a random event
- an outcome of interest
- an outcome that occurs most frequently
In the context of a Binomial Distribution, a "success" is defined as an outcome of interest, which could be positive, negative, or neutral depending on the context.
Which type of correlation is based on ranks and perfect for ordinal data?
- Kendall's Tau
- Pearson's correlation
- Point-Biserial Correlation
- Spearman's correlation
Spearman's correlation, also known as Spearman's rank correlation, is based on ranks and is perfect for ordinal data. It assesses how well the relationship between two variables can be described using a monotonic function. It is less sensitive to outliers and non-linear relationships compared to Pearson's correlation.