VAC_Exploring Data Insights_Assessment

What are some common techniques used in Data Analysis?

• Regression analysis, clustering, classification, machine learning, statistical analysis.
• Field that involves the statistical processing and interpreting data to uncover insights
• Classification, regression, association, probabilistic prediction, statistics
• Statistics, analysis, visualization, reporting

What is Predictive analytics

• Use historical data over other data to answer any question.
• Predict future performance using past data analysis
• Predict the future based on diagnostic analysis.
• Synthesize big data, mathematical science, business rule, and machine learning

What is Diagnostics Analytics

• Use descriptive analysis data and diagnose
• Use predictive analysis and diagnose
• Use historical data over other data to answer any question or for the solution of any problem
• Use prescriptive analysis and descriptive analysis

Data Mining process involves the following

• Extract data from unstructured , structured database
• It is about extracting the vital and valuable information from the data.
• Transform unstructured data and identify useful information
• Feature engineering using structured data

Data Clean up is required for better analytics outcome due to …

• Machine speed requirement, memory space requirement
• To handle large scale dataset
• Examining for errors, values, and inconsistencies, imputation, and getting rid of outliers.
• Getting rid of outliers, missing information

Data Analysis can be defined as follows

• Series of steps performed on a dataset for corporate report generation
• Performed on data set for academic research and study, prediction
• Series of step like Business Case Evaluation, Data Identification, Data Acquisition & Filtering, Data Extraction, Data Validation & Cleansing, Data Aggregation & Representation, Data Analysis, Data Visualization, and Utilization of Analysis Result
• Series of process like data collection, cleaning, analysis, modeling and reporting

underfitting results overall prediction performance as follows

• Underfitting results in poor performance and more bias
• Model learns the detail and noise in the training data
• Underfitting has high variance and low bias
• Underfitting has accuracy is acceptable

Hyper parameter is basically for tuning the performance by

• Parameters selected as part of training process to optimize
• Parameters that are not set manually
• Parameters part of Cross validation process
• Parameters explicitly defined by the user to control the learning process

Difference between Classification and regression is as follows

• Used for correlations and grouping respectively
• Used for parametric parameters only
• Used for grouping and correlation analysis respectively
• Used in supervised and unsupervised learning

Association can be defined as follows

• To find an association rule between a and b
• Discovering interesting relation between variables in a large dataset
• Associating the independent variable to the data set
• Associating dependent variable to a validation data set