Applications Of Ai

The phrase “garbage in, garbage out” is a popular saying in data science and AI fields.
It means that if you feed poor-quality or irrelevant data into an AI system, the output will also be poor.
AI models learn patterns from the data they are given, so inaccurate, incomplete, or biased data leads to flawed predictions and results.
AI systems cannot automatically fix fundamentally bad data during training — good preprocessing is essential.
Choosing the right AI model is important too, but this phrase specifically focuses on the data quality, not the model selection.
Avoiding garbage data only during deployment is not enough — the problem starts at data collection and preparation.
Therefore, if the data collected is bad, the AI model will not perform well — this is what “garbage in, garbage out” means.
So, the correct answer is option (A).

RMSE stands for Root Mean Square Error, not Rough Mean Square Error.
It is a widely used metric to measure the accuracy of predictions made by a model.
RMSE calculates the square root of the average of squared differences between actual and predicted values.
A lower RMSE value indicates better model performance, as it means the predictions are closer to actual results.
It does not find the square root of any number alone; it specifically relates to prediction error.
Therefore, RMSE helps to determine the prediction accuracy of an AI or statistical model.
So, the correct answer is option (C).

Selecting the right analytic approach depends on what kind of question you want to answer with your data.
Different problems require different techniques:
1. Descriptive Analytics answers “What has happened?”.
Example: A company uses descriptive analytics to create monthly sales reports showing total revenue and units sold.
2. Diagnostic Analytics answers “Why did it happen?”.
Example: If sales drop, diagnostic analytics finds the cause, such as seasonal trends or poor marketing.
3. Predictive Analytics answers “What is likely to happen?”.
Example: Using historical data to predict future customer demand for inventory planning.
4. Prescriptive Analytics answers “What should we do about it?”.
Example: An airline uses prescriptive analytics to optimize flight pricing and seat allocation.
By matching the analytic approach to the question, organizations can extract meaningful insights and make better decisions.
Choosing the wrong approach may waste time and lead to inaccurate conclusions.

When evaluating AI development platforms, it is important to check if they support open languages and offer features that enhance developer productivity.
(a) Open Languages:
Examples of popular open languages used in AI development include:
1. Python — An open-source, versatile language widely used for machine learning, deep learning, and data science tasks.
2. R — An open-source statistical computing language often used for data analysis and visualization in AI projects.
(b) Productivity Enhancing Capabilities:
These are features or tools that help developers build and deploy AI solutions more efficiently. Examples include:
1. Integrated Development Environment (IDE) Support — Built-in editors, debuggers, and testing tools that streamline coding.
For example: Jupyter Notebooks, PyCharm for Python.
2. Pre-built Libraries and Frameworks — Availability of ready-made packages such as TensorFlow, Scikit-learn, or Keras, which save time and effort.
Having open language support and productivity features makes an AI platform more flexible and user-friendly for developers.

Time series decomposition breaks down complex time-based data into simpler, interpretable parts.
Two key components are:
1. Trend — The trend component shows the long-term direction in the data over time.
It highlights whether values are generally increasing, decreasing, or remaining stable.
For example, a steady rise in yearly sales indicates an upward trend.
2. Seasonality — This component captures repeating patterns or cycles at regular intervals.
Seasonal effects can be monthly, quarterly, or annual.
For instance, higher ice cream sales during summer months show seasonality.
Other components include Level (average value) and Noise (random variations).
Together, these help in forecasting and understanding time-based patterns.

Train-Test Split Evaluation is a simple and quick technique used to measure the performance of a machine learning model.
In this method, the dataset is divided into two separate parts: training data and testing data.
The model is trained on the training set and then evaluated on the unseen test set to check how well it predicts new data.
This approach helps detect overfitting by comparing performance on training vs. test data.
Train-Test Split is easy to implement and works well for large datasets where splitting does not reduce data variety too much.
Two types of problems where Train-Test Split Evaluation can be used are:
1. Classification Problems — where the goal is to assign labels or categories to input data.
For example: spam detection, image classification.
2. Regression Problems — where the goal is to predict continuous numeric values.
For example: predicting house prices, temperature forecasting.
Train-Test Split gives a quick check of model performance but can be less reliable than cross-validation for small datasets.

Cross Validation is a robust statistical method used to assess how well a predictive model will perform on unseen data.
It helps detect overfitting and ensures that the model generalizes well.
The most common type is k-Fold Cross Validation.
In this procedure, the dataset is divided into k equal parts or folds.
The model is trained on (k-1) folds and tested on the remaining fold.
This process is repeated k times, with each fold used once as the test set.
The final performance score is the average of the scores from all k rounds.
Cross Validation provides a more reliable estimate of model performance compared to a single train-test split.
It is widely used for model selection, hyperparameter tuning, and validating predictive accuracy.

To understand and show relationships between variables in data, Statistical Analysis is the most suitable approach.
Statistical techniques like correlation analysis, regression, chi-square tests, and ANOVA are commonly used to identify and measure relationships.
These methods help uncover patterns such as how one variable affects another or how multiple factors interact.
Predictive and classification approaches focus more on predicting outcomes or categorizing data rather than directly showing relationships.
Descriptive analysis summarizes data and gives an overview but does not specifically test relationships in a robust statistical manner.
Therefore, to demonstrate and analyze relationships among data points, statistical analysis is the correct choice.
So, the correct answer is option (D).

Bias in selecting test data can seriously affect the results of an AI or machine learning project.
When humans choose test data with bias, it may not properly represent the real-world population or problem scenario.
This means the model will perform well on biased test data but fail in real-world applications where data varies more widely.
Bias can also lead to inaccurate measurement of the model’s performance, causing misleading conclusions.
An unbiased, representative test dataset ensures that the model is evaluated fairly for generalization and robustness.
Therefore, the statement is True — human biases during test data selection can indeed harm the testing phase.

Open-source frameworks like TensorFlow, Scikit-learn, and XGBoost are widely used tools for building and training AI and machine learning models.
TensorFlow is an open-source deep learning framework developed by Google for building neural networks and other complex AI tasks.
Scikit-learn is a popular Python library for traditional machine learning tasks such as classification, regression, and clustering.
XGBoost is an open-source implementation of gradient boosting algorithms, highly effective for structured data tasks and competitions.
On the other hand, Apple Siri is not an open-source framework — it is a commercial virtual assistant product developed by Apple.
Siri uses AI and natural language processing to understand and respond to user commands, but it is not a framework for developers to build AI models.
Therefore, the correct answer is option (B) Apple Siri.

AI models and machine learning techniques rely on finding patterns and relationships within data.
They learn these patterns to make predictions, classifications, or generate insights.
If there is truly no pattern in the data — for example, if the data is completely random or purely noise — then no model can learn anything meaningful from it.
Without patterns, the AI has nothing to detect, optimize, or improve upon.
Using AI in such cases would only waste computational resources and lead to unreliable or useless outputs.
Therefore, if there is no pattern in the data, AI development techniques should not be employed because they would have no meaningful basis for learning.
So, the given statement is False.

Assertion (A) says that MSE (Mean Squared Error) is a regression loss function.
This is true — MSE is widely used in regression tasks to measure the average squared difference between predicted and actual values.
It penalizes larger errors more severely due to squaring, making it a standard choice for regression models.
Reason (R) claims it is not good to use MSE if your target data is normally distributed around a mean value.
This is incorrect because MSE actually works well when data is normally distributed — in fact, the normal distribution assumption is one reason why MSE is commonly used.
If the target data is normally distributed, MSE provides an unbiased estimate of the prediction error.
Therefore, the assertion is true, but the reason given is false.
So, the correct option is (C) (A) is true but (R) is false.

Google Cloud AI Platform provides tools and services for building, training, and deploying AI models at scale.
IBM Watson Developer Platform offers a range of AI and machine learning APIs and services for developers.
Infosys Nia is an AI platform that provides cognitive capabilities and automation solutions for enterprises.
Alexa is Amazon’s virtual assistant, which uses AI and voice recognition to perform tasks.
However, Alexa itself is not an AI development platform — it is an application that runs on AI technologies built by Amazon.
While developers can build Alexa skills, it is not a general-purpose AI development suite like the other options.
So, the correct answer is option (D) Alexa AI.

Root Mean Square Error (RMSE) is a common metric used to measure the accuracy of a model’s predictions.
A lower RMSE indicates that the differences between predicted and actual values are smaller, meaning the model is more accurate.
Generally, the acceptable RMSE threshold depends on the dataset and context.
However, in most practical scenarios, an RMSE value lower than 180 is considered acceptable for a good-performing model.
If the RMSE is higher, it shows larger errors and suggests that the model may need improvement or tuning.
Model performance can be improved by better data preprocessing, parameter tuning, or selecting a better algorithm.
So, the correct answer is option (C) 180.

Step 1: Understand LLM risks.
LLMs are trained on large datasets, which may contain errors, biases, or sensitive information.

Step 2: Bias in training data.
If the training data contains social or cultural biases, the model may produce biased or unfair outputs.

Step 3: Data privacy issues.
Training data may include sensitive or personal information, leading to privacy risks if the model reproduces such data.

Step 4: Misinformation risk.
Models may generate incorrect or misleading information due to imperfect training data.

Step 5: Conclude risks.
Thus, major risks include bias, privacy concerns, and misinformation.

Step 1: Define Data Storytelling.
Data Storytelling is the process of presenting data in a meaningful and engaging way using visuals, narratives, and context to communicate insights clearly to the audience.

Step 2: Explain its components.
It combines three key elements: data, visuals (charts/graphs), and narrative (story) to make information easy to understand and impactful.

Step 3: State one reason for its importance.
One major reason why Data Storytelling has become powerful today is the huge increase in data generation, making it necessary to simplify complex data for better decision-making.

Step 4: Explain the impact.
It helps organizations and individuals quickly understand insights, identify trends, and take informed decisions effectively.

Step 1: Understanding Assertion.
Social media posts and images do not follow a structured format like tables, so they are examples of unstructured data, not structured data. Hence, Assertion (A) is false.

Step 2: Understanding Reason.
Unstructured data is defined as data that does not follow a predefined format. This statement is correct.

Step 3: Conclusion.
Thus, Assertion (A) is false, but Reason (R) is true.
Final Answer: (C)

Step 1: Understanding Generative AI.
Generative AI focuses on creating new content such as text, images, audio, or video based on patterns learned from training data.

Step 2: Evaluating options.
• (A): Classification task, not generative.
• (B): Related to classification boundaries.
• (C): Correct. Generative AI creates new data similar to training data.
• (D): Data cleaning, not generation.
Step 3: Conclusion.
Thus, the primary objective of Generative AI is to generate new data resembling training samples.
Final Answer: To generate new data that resembles its training samples.