Data Attributes and Missing Values Quiz

Categorical data where values represent distinct groups or states with no inherent order. Examples include hair color, marital status, and zip codes.

A nominal attribute with only two possible states, typically represented by 0 and 1. Examples include gender or a medical test result (positive/negative).

A type of attribute where the order of values matters, but the difference between them is not defined. Examples include grades (A, B, C), army rankings, or size categories (small, medium, large).

A type of attribute with equal interval sizes where values have order, but there's no true zero point. Examples include temperature in Celsius or Fahrenheit, and calendar dates.

A type of attribute with a true zero point, allowing for ratio comparisons. Examples include temperature in Kelvin, length, counts, and monetary quantities.

Random errors or variations in measured data.

Incorrect attribute values can be caused by faulty data collection tools, data entry mistakes, transmission errors, or limitations in technology.

Replacing missing values with a globally constant value, such as 'unknown' or a new class.

The practice of grouping data into bins to reduce the number of unique values and smooth out irregularities.

A technique for handling noisy data by smoothing values within bins using means, medians, or bin boundaries.

A mathematical way to measure the similarity between two vectors d1 and d2, often representing term-frequency vectors. It is calculated by taking the dot product of the two vectors and dividing by the product of their magnitudes.

The process of transforming raw data into a format suitable for analysis. It often involves cleaning, transforming, and reducing data, enabling improved accuracy and meaningful insights.

A type of data preprocessing that involves identifying and correcting errors, inconsistencies, or missing values within the dataset. It aims to ensure data quality and reliability for further analysis.

The process of combining data from multiple sources into a unified dataset. Integration addresses inconsistencies, duplicates, and data anomalies, creating a cohesive and consistent view.

A technique that uses a vector space model to represent words or documents as mathematical vectors, allowing for computational comparison of their semantic similarity. It involves measuring the angle between these vectors in a multi-dimensional space.

The limitations of traditional vector space models that often fail to capture the nuanced meanings and complexities of language. They are often unable to differentiate between words with similar surface structures but distinct semantic interpretations.

A technique that emphasizes the semantic relationships between words, going beyond simple word counts. It creates distributed representations of words, capturing the contextual and relational aspects of language.

A field in machine learning focused on learning meaningful representations of data, typically in the form of vectors. It aims to extract and encode the essence of information from complex data structures, enabling more accurate and efficient analysis.

A technique used in data integration to address situations where the same attribute or object has different names across multiple databases. The goal is to ensure consistency and avoid duplication.

A situation where an attribute can be derived from other attributes in a table, creating redundancy. For example, calculating age from date of birth.

The problem arising from redundant data, potentially leading to inconsistencies, storage inefficiencies, and difficulties in data analysis. Redundant attributes may be detected by correlation or covariance analysis.

A data transformation technique that aims to reduce the range of values in an attribute by scaling them to a smaller, specified range. Common methods include min-max normalization, z-score normalization, and normalization by decimal scaling.

A data transformation method that converts continuous data into discrete intervals. This can help simplify analysis, improve visualization, and reduce noise.

The process of reducing the size of a dataset without losing significant information. Techniques include lossy compression, which removes some data, and lossless compression, which maintains all data.

A data transformation approach that involves selecting a subset of data from a larger dataset, typically to reduce the size of the data or to make analysis more manageable.

A function that maps a set of attribute values to a new set of values, preserving the relationship between original and transformed values. This can involve various techniques like smoothing, attribute construction, aggregation, normalization, or discretization.

A graph that shows the relationship between two variables by plotting pairs of values as points in a plane. Each point represents a data pair, and the position of each point on the plane indicates the values of the two variables.

A method for graphically comparing the distributions of two datasets by plotting their quantiles against each other. It helps visualize if the two datasets follow similar distributions or deviate significantly.

A visual representation of data distribution using a box and whiskers. It shows the median, quartiles (Q1, Q3), and the range of the data, along with potential outliers.

A statistical measure that represents the difference between the third quartile (Q3) and the first quartile (Q1). It indicates the spread of the middle 50% of the data.

A set of five key statistical measures that summarize the distribution of data. It includes the minimum, first quartile (Q1), median, third quartile (Q3), and the maximum.

Values that fall outside a specific range, typically 1.5 times the interquartile range (IQR) below Q1 or above Q3. These values are often considered unusual or extreme.

A type of variable where the order of values is meaningful, but the differences between them may not be equal. Examples include ranks, grades, and opinion scales.

A way to represent documents or other objects as vectors, where each element (attribute) corresponds to the frequency of a specific feature (like a word or keyword) in that object.

Data that shows a positive relationship between two variables. As one variable increases, the other also tends to increase.

A data preprocessing technique where the data is divided into equal-sized intervals or bins. Each bin contains roughly the same number of data points.

A data smoothing technique where the values within a bin are replaced with a single representative value, like the bin's mean or boundary. This reduces noise and makes the data smoother.

A data smoothing technique where the values within a bin are replaced with the closest bin boundary value. This is another way to reduce noise and smooth the data.

A type of data compression where the original data can be fully reconstructed from the compressed data. This is a lossless compression method, meaning no information is lost.

A type of data compression where the original data cannot be fully reconstructed from the compressed data. This is a lossy compression method because some information is lost during compression.

A technique used in data compression that breaks down the signal into smaller fragments and reconstructs the original signal based on these fragments. It is often efficient but can also lead to quality loss.

In audio/video compression, data is often compressed in fragments without reconstructing the whole signal. Smaller parts can be retrieved for processing without the need for rebuilding the entire file.