Data Science Internship In Chennai

Femtosoft Technologies Best DataScience Internship In Chennai is designed for students and aspiring data scientists looking to gain real-world experience in data analysis, machine learning, and AI applications. This internship provides hands-on exposure to working with large datasets, building predictive models, and applying statistical techniques to solve business problems.

Note: After the timer finishes counting down, the exam will be submitted automatically

📝 Online Exam Instructions – Read Carefully Before Starting

1️⃣ Exam Details:

✅ Total Questions: (25 Questions)
✅ Exam Duration: (10 Minutes)

✅ Exam Type: MCQs

2️⃣ System Requirements:

✔ Use a laptop/desktop with a stable internet connection.
✔ Recommended browsers: Chrome, Firefox, Edge (latest versions).
✔ Ensure your webcam & microphone are working (if required).

3️⃣ Before Starting the Exam:

🖊 Fill in the Exam Registration Form Correctly

- - Full Name (This will be used for the certificate)
  - WhatsApp Number (Certificate will be sent to this number)
  - Email ID (Ensure it's correct; certificate & results will be emailed)

⚠️ Incorrect details may result in delays or disqualification.

4️⃣ Submission & Certification Process:

📌 Ensure all answers are submitted before the timer ends.
📌 Once submitted, you cannot change your responses.
📌 After passing, your certificate will be automatically generated with your registered name.

🎉 Congratulations on your success!

🎯 Stay focused and give your best! Good luck!

The number of attempts remaining is 2

Please fill all the required fields carefully, information provided will be used in offer letter and Completion Certificate.

1 / 25

1. Describe the impact of class imbalance on classification tasks.

A. B) It has no impact

B. D) None of the above

C. C) Only affects model accuracy

D. A) Leads to biased classifiers

2 / 25

2. What is the use of the train_test_split function in Python's scikit-learn library?

A. B) To merge datasets

B. A) To split data into training and testing sets

C. D) None of the above

D. C) It is not used in data preparation

3 / 25

3. What is the purpose of the train_test_split function?

A. B) To combine datasets

B. D) None of the above

C. A) To separate data into training and testing sets

D. C) It is not necessary for small datasets

4 / 25

4. What is a decision tree?

A. B) A type of linear regression

B. C) A clustering method

C. A) A flowchart-like structure that models decisions

D. D) None of the above

5 / 25

5. Explain the term "data wrangling."

A. A) The process of cleaning and unifying messy data

B. C) Only used for data analysis

C. B) A method of data storage

D. D) None of the above

6 / 25

6. Define the term "feature engineering."

A. C) Setting hyperparameters

B. B) A method of model evaluation

C. D) None of the above

D. A) The process of creating features to improve model performance

7 / 25

7. How does the k-nearest neighbors (KNN) algorithm work?

A. B) Uses weights based on distance

B. C) It gathers all classes in the model

C. D) None of the above

D. A) Assigns the majority class among K nearest samples

8 / 25

8. What is the significance of the ROC curve in model evaluation?

A. D) None of the above

B. B) It measures regression error

C. C) It only applies to clustering

D. A) It illustrates the trade-off between true positive rate and false positive rate

9 / 25

9. What is the purpose of the learning curve in machine learning?

A. C) To determine feature importance

B. D) None of the above

C. B) It's not useful

D. A) To show the relationship between training and validation performance as a function of training iterations

10 / 25

10. Describe the difference between L1 and L2 regularization.

A. D) None of the above

B. A) L1 shrinks coefficients to zero; L2 minimizes weights evenly

C. C) Both A and B

D. B) L1 minimizes weights linearly

11 / 25

11. What is K-means clustering?

A. C) A supervised learning algorithm

B. B) A decision tree method

C. D) None of the above

D. A) An algorithm that partitions n observations into k clusters

12 / 25

12. How can you handle missing data in a dataset?

A. D) None of the above

B. A) Imputation, deletion, or using algorithms that support missing values

C. C) Creating additional variables

D. B) Ignoring the entries

13 / 25

13. How do you choose the number of clusters in K-means?

A. C) Use the mean of the dataset

B. B) It’s arbitrary

C. D) None of the above

D. A) Using methods like the elbow method

14 / 25

14. What is a dataset's "sample"?

A. A) A single instance of data

B. C) The proportion of data used for training

C. B) The entire data collection

D. D) None of the above

15 / 25

15. How can you handle imbalanced datasets?

A. C) Increased cost of training

B. A) Using techniques like oversampling or undersampling

C. D) None of the above

D. B) Ignoring the minority class

16 / 25

16. What is the purpose of descriptive statistics?

A. C) To predict future outcomes

B. B) To analyze relationships between variables

C. A) To summarize and describe the main features of a dataset

D. D) None of the above

17 / 25

17. What is the role of natural language processing (NLP) in AI?

A. D) None of the above

B. A) To enable computers to understand and process human language

C. C) Both A and B

D. B) To enhance image recognition

18 / 25

18. What is the purpose of cross-validation?

A. B) To increase training accuracy

B. D) None of the above

C. C) Both A and B

D. A) To assess how the results of a statistical analysis will generalize to an independent data set

19 / 25

19. What are the common types of activation functions in a neural network?

A. B) Logarithmic, Exponential

B. A) Sigmoid, Tanh, ReLU

C. C) Polynomial

D. D) None of the above

20 / 25

20. What is the difference between bagging and boosting?

A. C) Both A and B

B. D) None of the above

C. B) Both methods improve accuracy

D. A) Bagging uses parallel models; boosting uses sequential

21 / 25

21. What are some common challenges faced during model deployment?

A. D) None of the above

B. A) Integration, scalability, and monitoring

C. C) Only trained models

D. B) Data collection issues

22 / 25

22. What is the maximum depth of a tree in decision tree algorithms?

A. D) None of the above

B. C) Both A and B

C. A) The maximum number of edges from the root to the leaf

D. B) The number of nodes

23 / 25

23. What is the purpose of dropout in neural networks?

A. A) To prevent overfitting by randomly ignoring some nodes during training

B. B) To initialize weights

C. C) It has no effect on neural network performance

D. D) None of the above

24 / 25

24. What is cross-entropy loss?

A. A) A measure of how well a model's predicted probability aligns with actual labels

B. D) None of the above

C. B) Used for regression problems

D. C) Used exclusively for clustering

25 / 25

25. What are embeddings in the context of natural language processing?

A. A) Continuous vector representations of words

B. D) None of the above

C. B) Only used in classification models

D. C) Statistical measures

AI & DATASCIENCE INTENSHIP

Data Science Internship In Chennai

📝 Online Exam Instructions – Read Carefully Before Starting

1️⃣ Exam Details:

2️⃣ System Requirements:

3️⃣ Before Starting the Exam:

4️⃣ Submission & Certification Process: