Course Outline
Business data analytics logic
1.1 Universal use of data
1.2 Two examples - hurricanes and customer behavior
1.3 Data science, engineering and data-driven decision making
1.4 Data processing and "Big Data"
1.5 From Big Data 1.0 to Big Data 2.0
1.6 Data and data analytics as strategic assets
1.7 Data analytics logic - summary
Business problems and solutions using data science
2.1 From a business problem to data mining
2.2 Supervised and unsupervised methods
2.3 Data Mining and Results
2.4 Consequences of managing data science projects
2.5 Analytical techniques and technologies
2.6 Summary
Predictive modeling - from correlation to supervised segmentation
3.1 Models, induction and forecasting
3.2 Supervised Segmentation
3.3 Visualization of results
3.4 Trees as sets of rules
3.5 Probability estimation
3.6 Case Study
3.7 Summary
Fitting the model to the data
4.1 Classification using mathematical functions
4.2 Regression
4.3 Class probability estimation and logistic "regression".
4.4 Non-linear functions
4.5 Neural networks
4.6 Summary
Overfitting and how to avoid it
5.1 Generalization
5.2 Overfitting
5.3 Analysis of the overfitting problem
5.4 Examples
5.5 Techniques for avoiding overfitting
5.6 Learning curves
5.7 Complexity check
5.8 Summary
Similarity, proximity and clusters
6.1 Similarity and distance measure
6.2 Nearest Proximity and Inference Rules
6.3 Key Techniques
6.4 Cluster analysis
6.5 Uses in Solving Business Problems
When is a model good?
7.1 Classifiers used in model evaluation
7.2 Generalizations beyond the limits of classification
7.3 Analytics Framework
7.4 Examples of application of basic evaluation techniques
7.5 Summary
Model visualization
8.1 Application of ranks
8.2 Profit Curves
8.3 Krzywe i grafy ROC (Receiver Operating Characteristics)
8.4 Area under the ROC curve
8.5 Cumulative Response
8.6 Examples
8.7 Summary
Evidence and probabilities
9.1 Example - Customer Focus
9.2 Probabilistic combinations of evidence
9.3 Application of Bayesian rules
9.4 Building a model
9.5 An example of using the model
9.6 Summary
Representing and exploring text
10.1 Why is text important?
10.2 Why is it difficult to work with text?
10.3 Representation
10.4 Example
10.5 Entropy in text
10.6 This is not a bag of words
10.7 Message Mining
10.8 Summary
Analytical engineering - case studies
Other tasks and techniques
12.1 Co-occurrences and Associations
12.2 Profiling
12.3 Relationship Forecasting
12.4 Information Reduction and Selection
12.5 Falsehoods, distortions and variance
12.6 Case studies
12.7 Summary
Business strategy and data science
13.1 Redux
13.2 Achieving a competitive advantage
13.3 Maintaining an Advantage
13.4 Resource Acquisition
13.5 New ideas and development
13.6 Maturity of the organization
How to conduct reviews of data science projects
End
Requirements
Training Data Science in business is addressed to several groups of people. Firstly, it is addressed to people from the business itself. Those who will work with statisticians and data analysts (data scientists, or as it is sometimes called in Poland "data masters"). Very often, these people will manage projects focused on business data analytics or will invest in data science projects. In addition to this group, the training is intended for those who will implement solutions focused on data analytics. In the case of these people, it is about presenting a platform for mutual understanding with business, which is not very interested in the details of the implementation itself. Of course, the third group should not be forgotten. About those who aspire to become a data master.
The training is not training in algorithms. It is also not a training in specific big data systems. Separate training courses are devoted to these topics, but without knowledge of certain fundamental concepts and principles of data science, projects in the field of data science are doomed to failure. Because the development of technology is very fast, it very often obscures the foundations on which solutions should be built, which the business can effectively use.
The training does not require sophisticated, specialist knowledge of statistics. Of course, you should be aware that, by its very nature, the material presented during the training is somewhat technical in nature. The aim of the training is to enable participants to gain a meaningful understanding of data science, not just a general overview of the field. Despite this rather ambitious goal, the mathematical apparatus is limited to the absolutely necessary minimum. Generally speaking, the training includes everything you need to understand to design and build advanced, data science-based solutions to business problems.
Testimonials (1)
duża wiedza, praca na rzeczywistych danych