An example of Predictive Analytics

Predictive analytics uses data to forecast future outcomes

What is Predictive Analytics

Forecasting and predicting trends using historical data

The benefits of Predictive Analytics

Improves decision making as the business can act proactively instead of reactively, allowing you to be better prepared ahead of time.

What competitive advantage does it provide?

Gives a competitive edge over your competition if you can identify areas of weakness or trends over others.

The biggest risk with Predictive Analytics.

Data quality is a risk, if the quality of your data isn’t consistent or accurate, you may be making incorrect predictions.

Additional points on the risk of using Predictive Analytics

Over reliant on predictive models rather than your taking a step back and truly analysing the data yourself and taking your own approach. We sometimes know industries better than machines.

How would you ensure source data quality?

Clean the data; Remove errors; Standardise the data; Remove or fill missing values; Keep data uptodate; Check for bias; Check meta data of a file; Check file hashtags for data transfers

How can you check for the authenticity of a source data?

Check to find the root source of the data; Check the data lineage or audit trail of where the data came from; cross reference sources with other searches and authors for trusted confirmations of data sources.

Statistical Sanity Testing using Benford's Law

In many naturally occurring datasets (like financial records or street addresses), the leading digit is more likely to be a 1 (about 30% of the time) than a 9 (less than 5%). If the leading digits are evenly distributed (11% for each), the data may be fabricated.

Genuine data usually has a natural amount of "noise." An identical row appearing thousands of times without a logical reason (like a system error) suggests the data was "padded."

Help

Options

focusNode

Didn't know it?
click below

Knew it?
click below

Don't Know

Remaining cards (0)

Know

retry

shuffle

restart

0:00

Embed Code - If you would like this activity on your web page, copy the script below and paste it into your web page.

Normal Size Small Size show me how

DA - K14-S11

“The principles of descriptive predictive and prescriptive analytics.

Term	Definition
An example of Predictive Analytics	Predictive analytics uses data to forecast future outcomes
What is Predictive Analytics	Forecasting and predicting trends using historical data
The benefits of Predictive Analytics	Improves decision making as the business can act proactively instead of reactively, allowing you to be better prepared ahead of time.
What competitive advantage does it provide?	Gives a competitive edge over your competition if you can identify areas of weakness or trends over others.
The biggest risk with Predictive Analytics.	Data quality is a risk, if the quality of your data isn’t consistent or accurate, you may be making incorrect predictions.
Additional points on the risk of using Predictive Analytics	Over reliant on predictive models rather than your taking a step back and truly analysing the data yourself and taking your own approach. We sometimes know industries better than machines.
How would you ensure source data quality?	Clean the data; Remove errors; Standardise the data; Remove or fill missing values; Keep data uptodate; Check for bias; Check meta data of a file; Check file hashtags for data transfers
How can you check for the authenticity of a source data?	Check to find the root source of the data; Check the data lineage or audit trail of where the data came from; cross reference sources with other searches and authors for trusted confirmations of data sources.
Statistical Sanity Testing using Benford's Law	In many naturally occurring datasets (like financial records or street addresses), the leading digit is more likely to be a 1 (about 30% of the time) than a 9 (less than 5%). If the leading digits are evenly distributed (11% for each), the data may be fabricated.
Duplicate Analysis	Genuine data usually has a natural amount of "noise." An identical row appearing thousands of times without a logical reason (like a system error) suggests the data was "padded."

Created by: PRO Teacher Adamrsc

"Know" box contains:
Time elapsed:
Retries: