Q: What is PCA (Principal Component Analysis)?

**PCA** is a **Dimensionality Reduction** technique. It transforms a large set of variables into a smaller one that still contains most of the information in the large set. * It creates new variables (**Principal Components**) which are linear combinations of the original variables. * The 1st PC accounts for the maximum variance in the data. * **Use Case**: Reducing computation time and visualization (3D -> 2D).

Question 1

Describe the Data Science Lifecycle.

Accepted Answer

The lifecycle generally consists of:
1.  **Business Understanding**: Asking relevance questions and defining objectives.
2.  **Data Mining**: Gathering and scraping data.
3.  **Data Cleaning**: Fixing inconsistencies and handling missing values.
4.  **Exploration (EDA)**: Forming hypotheses and visualizing patterns.
5.  **Feature Engineering**: Selecting and transforming variables.
6.  **Predictive Modeling**: Training machine learning models (splitting data into train/test).
7.  **Data Visualization**: Communicating results to stakeholders.

Question 2

How do you handle Missing Data?

Accepted Answer

1.  **Deletion**:
    *   *Listwise*: Remove entire row (if dataset is huge).
    *   *Pairwise*: Remove only when specific variable is needed.
2.  **Imputation**:
    *   **Mean/Median/Mode**: Fill with average (can reduce variance).
    *   **Forward/Backward Fill**: For time-series data.
    *   **Prediction**: Use a model (KNN or Regression) to predict missing values based on other features (most accurate).

Question 3

Pandas Series vs DataFrame?

Accepted Answer

### Series
*   A **one-dimensional** labeled array capable of holding any data type.
*   Like a single column in an Excel sheet.

### DataFrame
*   A **two-dimensional** labeled data structure with columns of potentially different types.
*   Like the entire Excel sheet or SQL table.
*   Technically, a DataFrame is a collection of Series objects sharing the same index.

Question 4

Correlation vs Causation?

Accepted Answer

**"Correlation does not imply Causation."**

*   **Correlation**: A statistical measure (Person's r) that indicates the extent to which two or more variables fluctuate together. (-1 to +1).
*   **Causation**: Indicates that one event is the result of the occurrence of the other event.

*Example*: Ice cream sales and shark attacks are positively correlated (both increase in summer), but eating ice cream does not cause shark attacks.

Question 5

What is the Central Limit Theorem (CLT)?

Accepted Answer

The **CLT** states that the sampling distribution of the sample mean will approach a **Normal Distribution** (Bell Curve) as the sample size gets larger, regardless of the shape of the population distribution.

*   *Significance*: This justifies the use of parametric hypothesis tests (like t-tests) even on non-normal data, provided the sample size is sufficient (usually n > 30).

Question 6

What is a P-Value?

Accepted Answer

The **p-value** is the probability of observing results at least as extreme as those measured when the **Null Hypothesis ($H_0$)** is true.

*   **Low p-value (< 0.05)**: Strong evidence against $H_0$. We reject the Null Hypothesis. (Statistically Significant).
*   **High p-value**: Weak evidence against $H_0$. We fail to reject it.

Question 7

What is PCA (Principal Component Analysis)?

Accepted Answer

**PCA** is a **Dimensionality Reduction** technique.

It transforms a large set of variables into a smaller one that still contains most of the information in the large set.
*   It creates new variables (**Principal Components**) which are linear combinations of the original variables.
*   The 1st PC accounts for the maximum variance in the data.
*   **Use Case**: Reducing computation time and visualization (3D -> 2D).

Question 8

Explain Cross-Validation.

Accepted Answer

**Cross-Validation** is a resampling procedure used to evaluate machine learning models on a limited data sample.

**K-Fold Cross-Validation**:
1.  Shuffle the dataset randomly.
2.  Split the dataset into **k** groups.
3.  For each unique group:
    *   Take the group as a hold out or test data set.
    *   Take the remaining groups as a training data set.
    *   Fit a model on the training set and evaluate it on the test set.
4.  Summarize the skill of the model using the sample of model evaluation scores.

It gives a less biased estimate of model performance than a simple train/test split.

Question 9

What is Exploratory Data Analysis (EDA)?

Accepted Answer

**EDA** is an approach to analyzing data sets to summarize their main characteristics, often with visual methods.

### Goals:
*   Spot anomalies and outliers.
*   Test hypotheses.
*   Check assumptions required for model fitting.
*   Discover patterns.

*Tools*: Histograms, Box plots, Scatter plots, Heatmaps.

Question 10

Inner Join vs Left Join (SQL)?

Accepted Answer

### Inner Join
Returns records that have matching values in **both** tables.
*   intersection of sets.

### Left Join (Left Outer Join)
Returns all records from the **left** table, and the matched records from the right table.
*   The result is NULL from the right side, if there is no match.

Data Science

1Describe the Data Science Lifecycle.

2How do you handle Missing Data?

3Pandas Series vs DataFrame?

4Correlation vs Causation?

5What is the Central Limit Theorem (CLT)?

6What is a P-Value?

7What is PCA (Principal Component Analysis)?

8Explain Cross-Validation.

9What is Exploratory Data Analysis (EDA)?

10Inner Join vs Left Join (SQL)?