Women’s Week Challenge - Day 2

[Advika_]

Ready for another round of brain teasers and AI fun? Let's go! 🎯

Today's Challenges:

🧠 If you could teach AI one human emotion, what would it be and why?

🧠 A machine learning model has the following metrics: Precision = 80%, Recall = 70%, Total Predictions = 200, Positive Predictions = 120. Now, can you solve these?
➡️ If you randomly select a positive prediction, what is the probability that it is actually correct?
➡️ (Optional) How many actual positive cases exist in the dataset?

🧠 What's a common Data Engineering mistake in Databricks that could slow down pipelines?

⁉️ Unique Mystery Challenge - The Case of the Missing Data!
Your ETL pipeline runs successfully, but the output table is mysteriously empty. What could have gone wrong? List possible reasons and help debug the mystery!

How to Play: Drop your answers below and engage!

Missed the full details? 
Visit: Women's Week Challenge: Play, Engage & Win Swag 🔗

[seohunter106]

😊

[Nikita_Garg]

1. Empathy; AI interacting with humans in customer support or healthcare could provide better responses by understanding emotions. It would reduce frustration and make AI feel more Human.

2. 75%??

3. Failing to run OPTIMIZE and VACUUM leads to slowing read/write operations.

4. Null values, if key columns are NULL.

[Shibin_P]

🧠 If you could teach AI one human emotion, what would it be and why?

Loyalty, It will be good to teach loyalty to AI so it doesn't switch sides based on the data. AI should value the connections and care for those people with whom it interacts for a long time

🧠 A machine learning model has the following metrics: Precision = 80%, Recall = 70%, Total Predictions = 200, Positive Predictions = 120. Now, can you solve these?
➡️ If you randomly select a positive prediction, what is the probability that it is actually correct?
➡️ (Optional) How many actual positive cases exist in the dataset?

I think, since precision is the probability of actually correct predictions, the answer should be 80%. Am I correct?

🧠 What's a common Data Engineering mistake in Databricks that could slow down pipelines?

Usage of display() instead of show() for large data frames.

⁉️ Unique Mystery Challenge - The Case of the Missing Data!
Your ETL pipeline runs successfully, but the output table is mysteriously empty. What could have gone wrong? List possible reasons and help debug the mystery!

• If the pipeline is connecting to the wrong source or wrong credentials 
• The query used to fetch data is incorrect

[kajal_adik]

1. Regret, AI should understand and learn from the impact of bad decisions. It would help improve ethical decision-making in AI applications.

2. Probability of a correct positive prediction: 80%?

3. Some users repeatedly query large datasets without using CACHE TABLE, leading to unnecessary recomputation and performance slowdown.

4. A WHERE clause unintentionally excluded all records or the Source schema doesn’t align with the target table.

[Aadi13]

1) Teaching AI about happiness would allow it to create more positive interactions. If AI understands happiness, it could encourage people during difficult times.

2) 80%

3) A common Databricks mistake is inefficient partitioning in Delta Lake tables, which slows pipelines. Over-partitioning increases I/O overhead, while under-partitioning forces full-table scans.

4) Lack of Write Permissions