Data Mining Techniques

 Data Mining Techniques

I have previously spoken about data analysis and the real world impact of it. Today, I discovered a website which highlighted the common data mining techniques which are used to discover hidden patterns and correlations is vast amounts of data, which is then used for decision making and predictive modelling.

Some of the most common data mining techniques I discovered are:

  • Classification - A technique used to categorise data into predefined classes based on the data's features.

  • Association Rule - Discovering interesting relationships or patterns among a set of items in transactional or market basket data. It helps identify frequently co-occurring items and generates rules to show associations between items, such as if  X, then Y.

  • Anomaly detection - Aims to identify unusual data instances that deviate significantly from the expected patterns. Some use cases may be detecting fraudulent actions or network intrusions.

  • Clustering - To group similar data instances together based on their similarities to discover natural patterns or structures in the data without predefined classes.

  • Time series analysis - analysing and predicting data points collected over time.

  • Regression - To predict numeric values based on the relationship between input variables and a target variable to find a mathematical function or model that best fits the data to create accurate predictions from it.



https://www.qlik.com/us/data-analytics/data-mining






Comments

Post a Comment

Popular posts from this blog

Limitations of Traditional Statistics

Image
 Limitations of Traditional Statistics In my last post, I spoke about the use of traditional statistics. Now, I will cover some of the drawbacks of using traditional statistics to make predictions in the real world. Traditional data analysis has been practised with structured data, often in a relational database; however, Big Data isn’t as manageable with this structured format. Traditional statistics is about analysing and summarising data and is suited to processing smaller amounts of linear, repeatable data to draw a conclusion. The easiest environments are when the data has a stable relationship – e.g. regular exercise increases, obesity decreases. Therefore, the practicality of traditional data analysis is limited when applied to the real world. More recently, there is now the availability of new large data sets and cheaper and faster computing power. In addition, there has been a surge in machine learning research and development, which aims to automatically discover re...
Read more

Application of Big Data Techniques to a Problem

 Application of Big Data Techniques to a Problem From the previous posts, it is clear that Big Data is collected, stored and utilised in the real world at enormous scales by huge companies. Now I want to show directly how these companies use this data to address their problems in the modern world. Netflix Netflix is known for its great, vast choice of shows that users love. But how does Netflix know what each individual customer might want to watch? Netflix recommends specific titles for each user on the home page. It uses multiple algorithms for personalised rankings, searches, similarity ratings, ranking and more to narrow down the shows to display on the recommended screen for each individual customer. It looks at what titles the user has watched, searched for, and rated good or bad so the system can show them similar content. Additionally, if user A shows the same interests as user B, Netflix will show user A the user B's recommended shows. Each show is also put into some of th...
Read more

Traditional Statistics

 Traditional Statistics Today we learned about types of traditional statistics and how they're used to draw conclusions.  Statistics is the discipline that concerns the collection, organisation, analysis, interpretation, and presentation of data. It can be broken up into descriptive and inferential statistics. Descriptive Statistics Only the subjects that have been analysed have findings implied upon them. If we had the political views of 20 people, we may be interested in the voting intentions of these people or the distribution or spread of the votes, but the findings are limited to only these 20 people. Inferential Statistics If you don’t have access to the whole population you’re interested in, but only a small portion. If you want to study the political voting intentions of everybody in Scotland. From what I've learnt about the two types, I would say that descriptive is more accurate as conclusions are drawn from what the data had been directly collected from. Howev...
Read more