Other Flashcards
What is an IOT (internet of things)
Physical, smart (without human intervention, data driven) and wireless. e.g. Smart Fridge.
Priority?
What, How, Why.
Data Metrics
Market Reputation, Customer Retention, Operational effeciency.
Data metrics are quantifiable measures used to track and assess the status of specific business processes. Essentially, they are a set of data points that are used to quantify performance, behaviors, and other relevant business activities. Metrics provide a way to measure the effectiveness, performance, and progress of a project or plan.
Here are a few key points about data metrics:
Specific: They are not vague indicators but are tied to specific data points that can be measured accurately.
Quantifiable: Metrics can be represented in numbers, making it easier to track changes and trends over time.
Standardized: Effective metrics have standardized definitions and are understood and measured consistently across an organization.
Relevant: They are directly related to business objectives, ensuring that the focus is on tracking the right aspects of business performance.
Actionable: Good metrics should inform decision-making and potentially lead to actions that can improve the measured activity.
Examples of data metrics include conversion rates in marketing, sales growth, customer satisfaction scores, production costs, and many others depending on the industry and the specific area of the business they are meant to measure.
KPI 5 requirements
Direction (increase or Decrease)
Metric
Baseline
Target
Timeframe
Impact Diagram
Circle - above target
Square - Below target
Diamond - neutral
Class Classification - Upsell/Cross Sell
This technique predicts which customers might buy more expensive items (upsell) or additional products (cross-sell) based on their purchase history. It helps tailor marketing to offer personalized product recommendations, boosting sales and customer experience.
2-Class Classification Question:
A 2-class classification question is a binary decision problem where a model determines whether an input belongs to one of two possible categories.
Difference between Data and Metrics
Raw Facts: Data is the raw, unprocessed facts that are collected from various sources.
Unstructured or Structured: It can be unstructured (like text or images) or structured (like numbers or dates).
Measurement: A metric is a quantifiable measure that is used to track and assess the status of a specific business process.
Derived from Data: It’s derived from data by applying some form of calculation or algorithm.
Performance Indicators: Metrics are often key performance indicators (KPIs) that help businesses understand how they are performing against their goals.
4 AI business Strategies
Efficiency - lower complexity - high data
Effectiveness - lower complexity - low data
Innovation - higher complexity - high data
Expert - higher complexity - low data
Data Mining
Drilling down the data.
Regression - Data mining
Predictive Data mining
The process of going through and system data bases and finding relevant data to analyse for the purpose of prediction. e.g algorithm-based tools to go through customer data base to look at past transactions to support future volumes of transactions.
Predictive Data Analytics
Using past data to forecast future events. Steps include:
Data Collection: Gather relevant past and current data.
Data Cleaning: Remove errors and irrelevant information.
Data Analysis: Identify patterns using statistical methods.
Model Building: Create models with algorithms to predict outcomes.
Validation: Test models for accuracy with a separate data set.
Deployment: Use the model in real-world scenarios to make predictions.
Update: Regularly refine the model with new data and methods.
Purpose: To anticipate trends for informed decision-making in fields like finance, marketing, healthcare, and operations.
Algorithm?
Clear starting - inputting data and stopping point - exporting data
