20+ Tiger Analytics Senior Analyst Interview Questions and Answers

The probability of getting 5 Sundays in a 31 day month is less than 1%.

• There are 7 days in a week, so the probability of any given day being a Sunday is 1/7.

• In a 31 day month, there are 4 full weeks and 3 extra days.

• The probability of the first 4 weeks having 4 Sundays is (1/7)^4.

• The probability of the remaining 3 days being Sundays is (3/7).

• Multiplying these probabilities gives a total probability of less than 1%.

Probability of seeing a plane in 30 minutes given 15% chance in 10 minutes.

• Calculate the probability of not seeing a plane in 10 minutes

• Use the formula P(X>=1) = 1 - P(X=0)

• Calculate the probability of not seeing a plane in 30 minutes using the above probability

• Calculate the probability of seeing atleast 1 plane in 30 minutes using the formula P(X>=1) = 1 - P(X=0)

Probability of a random point in a circle of 1 unit radius being closer to the circumference than the center.

• The probability is 1/4 or approximately 0.785.

• This is because the area of the circle closer to the circumference is 1/4th of the total area.

• This can be calculated using the formula for the area of a circle: A = πr^2.

Implementing Python Collection Counter from Scratch

• Create an empty dictionary to store the elements and their count

• Iterate through the input list and add elements to the dictionary with their count

• Return the dictionary

• Example: input_list = ['apple', 'banana', 'apple', 'orange', 'banana']

• Output: {'apple': 2, 'banana': 2, 'orange': 1}

Categorical features in Linear Regression require encoding using dummy variables. Removing one dummy variable avoids the dummy variable trap.

• Categorical features need to be encoded using dummy variables to be used in Linear Regression

• Dummy variable trap occurs when one dummy variable can be predicted from the others

• Removing one dummy variable avoids the issue of multicollinearity and improves model performance

• Example: Gender (Male/Female) can be encoded as a dummy variable wi...read more

Program to check if a number is power of 3

• Use logarithm to check if the result is an integer

• Check if the number is greater than 0

• Check if the remainder is 0 when the number is divided by 3 repeatedly

Matrix multiplication, factorial and Fibonacci series implementation

• Matrix multiplication involves multiplying two matrices to get a third matrix

• Factorial is the product of all positive integers up to a given number

• Fibonacci series is a sequence of numbers where each number is the sum of the two preceding ones

• Factorial can be implemented using recursion or iteration

• Fibonacci series can be implemented using recursion or iteration

Using Bayes Theorem, find the probability of getting 10 heads given 99 unbiased coins and 1 biased coin.

• Identify the prior probability of getting 10 heads with unbiased coins

• Calculate the likelihood of getting 10 heads with the biased coin

• Use Bayes Theorem to calculate the posterior probability of getting 10 heads given the mix of coins

• Consider the impact of the biased coin on the overall probability

Entropy is a measure of randomness or disorder in a system. Gini index is a measure of impurity in a dataset. Derivatives measure rate of change. P-value is the probability of observing a test statistic. Beta value is the coefficient in a regression model. Imbalanced datasets have unequal class distribution. Recall is the proportion of actual positives correctly identified. Precision is the proportion of predicted positives that are actually positive. Slope in one variable is...read more

Accuracy score can be misleading on imbalanced datasets.

• Accuracy score can be high even if the model is not performing well on the minority class.

• F1 score, precision, and recall are better metrics for imbalanced datasets.

• Stratified sampling, oversampling, and undersampling can help balance the dataset.

• Example: A model predicting cancer in a dataset with only 1% positive cases.

• Using accuracy score, a model that always predicts negative will have 99% accuracy.

• However, this mode...read more

Different regression models are used based on the type of data and relationship between variables.

• Linear regression is used when there is a linear relationship between the independent and dependent variables.

• Logistic regression is used when the dependent variable is binary.

• Polynomial regression is used when the relationship between variables is non-linear.

• Ridge regression is used when there is multicollinearity in the data.

• Lasso regression is used when feature selection is im...read more

Different varieties of Fibonacci series in Python.

• Standard Fibonacci series

• Fibonacci series with user-defined starting numbers

• Fibonacci series with user-defined length

• Fibonacci series with user-defined step

• Fibonacci series with user-defined function

I have used various ML algorithms such as linear regression, decision trees, random forests, and neural networks in my projects.

• Linear regression for predicting continuous values

• Decision trees for classification and regression tasks

• Random forests for ensemble learning and improved accuracy

• Neural networks for complex pattern recognition

To maximize profit, buy when the stock price is low and sell when it is high.

• Identify the lowest price point to buy the stock

• Identify the highest price point to sell the stock

• Consider market trends and analysis for optimal buying and selling days

P-value in regression summary measures the probability of observing a test statistic as extreme as the one computed from the sample data.

• P-value is used to determine the statistical significance of the regression coefficient.

• A low P-value (less than 0.05) indicates that the coefficient is statistically significant.

• A high P-value (greater than 0.05) indicates that the coefficient is not statistically significant.

• P-value is calculated using the t-test or F-test depending on the...read more

Sample T test is a statistical test used to determine if there is a significant difference between the means of two groups.

• It is used to compare the means of two groups.

• It assumes that the data is normally distributed.

• It is commonly used in research studies to determine if a treatment has a significant effect.

• Example: A sample T test can be used to compare the mean weight of two groups of people who followed different diets.

R-squared is a statistical measure that represents the proportion of the variance for a dependent variable that's explained by an independent variable.

• R-squared ranges from 0 to 1, with 1 indicating that all variance in the dependent variable is explained by the independent variable.

• It is used in regression analysis to determine how well the regression line fits the data points.

• A higher R-squared value indicates a better fit of the model to the data, while a lower value sugge...read more

The architecture diagram of the project showcases the overall structure and components of the system.

• The architecture diagram typically includes components like servers, databases, APIs, and client applications.

• It shows how these components interact with each other and the flow of data within the system.

• Commonly used tools for creating architecture diagrams include Microsoft Visio, Lucidchart, and draw.io.

Developed a data analysis project to optimize marketing strategies for a retail company.

• Utilized customer segmentation techniques to identify target demographics

• Analyzed sales data to determine the most effective marketing channels

• Implemented A/B testing to measure the impact of different marketing campaigns

MSE metric is commonly used in data analysis to measure the average squared difference between predicted values and actual values.

• MSE helps to quantify the accuracy of a model by penalizing large errors more than small errors.

• It is easy to interpret as it gives a clear measure of how well the model is performing.

• MSE is differentiable, making it suitable for optimization algorithms like gradient descent.

• Example: In linear regression, MSE is often used to evaluate the performan...read more

MSE metrics are commonly used to measure the average squared difference between predicted values and actual values in statistical analysis.

• MSE helps in evaluating the performance of a predictive model by quantifying the accuracy of the model's predictions.

• It penalizes large errors more heavily than small errors, making it a useful metric for identifying outliers or areas where the model is underperforming.

• MSE is widely used in machine learning, regression analysis, and time s...read more

Probability is the likelihood of a specific event occurring, expressed as a number between 0 and 1.

• Probability ranges from 0 (impossible event) to 1 (certain event)

• It can be calculated by dividing the number of favorable outcomes by the total number of possible outcomes

• Probability can be represented as a percentage, fraction, or decimal

Our tech stack includes Java, Spring Boot, Angular, and PostgreSQL.

• Java

• Spring Boot

• Angular

• PostgreSQL