# **Data Types and Introduction to Machine Learning** This section covers the fundamental types of data used in computer science and provides an introduction to the core concepts of Machine Learning, including model building, evaluation, and the distinction between different learning algorithms. ## **4.1 Data Types** Data is a collection of raw facts and figures. In data science, data is categorized into various types. | Data Type | Description | Example | | :--- | :--- | :--- | | **Numerical** | Quantitative data used to represent numerical values. | Marks, Salary, Age | | **Categorical** | Data with discrete values. It has two sub-types:
- *Nominal*: Categories without a specific order.
- *Ordinal*: Categories with a meaningful order. | *Nominal*: Male, Female, Color
*Ordinal*: Rank, High, Low, Small, Large | | **Boolean** | Data with only two possible values, typically 0 or 1. | True/False, Yes/No | | **Text Data** | A sequence of characters. | Description, Review, Feedback | | **Time Series** | Data collected over a period of time. | Stock price, weather data | | **Image** | Data represented as pixel values. | X-Ray, Ultrasound, Photographs | | **Audio** | Data in the form of sound waves. | Speech, Music | | **Structured** | Data organized in a fixed format, like rows and columns. | Spreadsheets, Databases | | **Semi-Structured** | Data organized in a specific format but without a fixed schema. | XML, JSON files | | **Unstructured** | Data without any predefined structure or organization. | Videos, Emails, Web Pages | ## **4.1.1 Machine Learning vs. Rule-based Algorithms** A fundamental distinction exists between systems that learn from data and those that follow explicit instructions. - ***Machine Learning (ML)***: Models learn patterns directly from data. They can adapt and improve their performance as they are exposed to more data. ML is suitable for complex problems where rules are not easily defined. - ***Rule-based Algorithms***: These systems operate based on a set of predefined, fixed instructions (rules). They are consistent but inflexible and require manual updates to handle new situations. | Aspect | Machine Learning | Rule-based Algorithms | | :--- | :--- | :--- | | **Advantages** | - Handles complex and unstructured data.
- Automatically learns patterns.
- Does not require explicit human expertise for pattern detection.
- Accuracy can improve with more data. | - Does not require large datasets.
- Performance is consistent and predictable.
- Simpler and requires fewer computational resources. | | **Limitations** | - Performs poorly with insufficient data.
- Decision-making can be a "black box," difficult to interpret.
- Requires significant computational resources.
- Prone to *overfitting* and *bias* from training data. | - Inflexible; cannot adapt to new, unanticipated data.
- Performs poorly with unstructured or ambiguous data.
- Cannot learn or improve from experience.
- Can become slow if too many rules accumulate. | ## **4.1.2 Model Building in Machine Learning** This is the process of creating a system that learns from data to make predictions. The key steps are: 1. **Data Collection & Preprocessing**: Gathering and cleaning data to make it suitable for analysis (e.g., handling missing values, normalizing data). 2. **Algorithm Selection**: Choosing a suitable ML algorithm (e.g., classification, regression) based on the problem. 3. **Model Training**: The algorithm learns from a *training dataset* by adjusting its internal parameters to minimize errors. 4. **Model Evaluation**: The model's performance is tested on a separate *validation dataset* to see how well it performs on unseen data. 5. **Testing**: The final, optimized model is assessed on a *test dataset* to measure its effectiveness. 6. **Deployment**: If satisfactory, the model is used in a real-world application. ### **Key Terms in Model Building** - ***Feature Engineering***: The process of selecting and transforming data features to improve a model's performance. For example, identifying features like eye shape, ear size, and color to differentiate between images of cats, dogs, and horses. - ***Test-Train Split***: Dividing the dataset into two parts: a larger portion for *training* the model and a smaller portion for *testing* its performance. A common split is 70% for training and 30% for testing. - ***Model Training***: The core process where the model learns from labeled or unlabeled data. The main types are: - *Supervised Learning*: Training on labeled data (e.g., images labeled 'cat' or 'dog'). *Linear Regression* is a common algorithm. - *Unsupervised Learning*: Training on unlabeled data to find hidden patterns. *K-Means Clustering* is a common algorithm. - *Reinforcement Learning*: The model learns by interacting with an environment and receiving rewards or penalties. *Q-Learning* is a popular algorithm. - ***Overfitting***: Occurs when a model learns the training data too well, including its noise. It performs excellently on training data but poorly on new, unseen data. - ***Underfitting***: Occurs when a model is too simple to capture the underlying patterns in the data. It performs poorly on both training and new data. - ***Model Assessment***: Evaluating a model's performance using specific metrics. ### **Model Evaluation Metrics** - **Confusion Matrix Terms**: - *True Positive (TP)*: Correctly predicted positive. (e.g., Model predicts "Cat," and it is a Cat). - *False Positive (FP)*: Incorrectly predicted positive. (e.g., Model predicts "Cat," but it is a Dog). - *True Negative (TN)*: Correctly predicted negative. (e.g., Model predicts "Not Cat," and it is a Dog). - *False Negative (FN)*: Incorrectly predicted negative. (e.g., Model predicts "Not Cat," but it is a Cat). - ***Accuracy***: The percentage of total predictions that are correct. $$ \text { Accuracy }=\frac{\text { Correct Predictions }}{\text { Total Predictions }} $$ - ***Precision***: Measures the accuracy of positive predictions. Of all the times the model predicted "positive," how many were actually positive? $$ \text { Precision }=\frac{\text { True Positives }}{\text { True Positives + False Positives }} $$ - ***Recall***: Measures the model's ability to find all actual positive cases. Of all the actual positive cases, how many did the model correctly identify? $$ \text { Recall }=\frac{\text { True Positives }}{\text { True Positives + False Negatives }} $$ - ***F1 Score***: The harmonic mean of Precision and Recall, providing a single score that balances both metrics. It is especially useful for imbalanced datasets. $$ F1_{\text {Score }}=2 \cdot \frac{\text { Precision } \cdot \text { Recall }}{\text { Precision }+ \text { Recall }} $$ ## **4.1.3 Learning from Data** Machine learning models improve their performance by identifying patterns, trends, and relationships within data. The quality and quantity of data are crucial; more high-quality data generally leads to better learning and more accurate predictions. This continuous learning process allows the model to refine its understanding and improve its accuracy over time. ## **4.1.4 Algorithms for Different Data Structures** Choosing the right algorithm is critical for building an effective ML model. The choice depends on the data type and the problem to be solved. - **Based on Data Structure**: - ***Structured Data*** (e.g., databases, CSV files): Traditional algorithms like *Regression* and *Classification* are suitable. - ***Unstructured Data*** (e.g., text, images, audio): Advanced algorithms like *Neural Networks* are better suited to handle the complexity and lack of a predefined format. - **Based on Learning Task**: - ***Regression***: Predicts continuous values (e.g., prices). - ***Classification***: Sorts data into predefined categories (e.g., spam vs. not spam). - ***Clustering***: Groups similar, unlabeled data points together. - ***Neural Networks***: Used for complex pattern recognition in large datasets, such as image recognition or natural language processing.