Algorithm / Algorithms
In a broader sense, algorithms are “step-by-step instructions” that are intended to lead to a specific goal. The word algorithm is mostly used in connection with mathematical objects. The individual steps of an algorithm are usually mathematical operations. Translating these operations into software code creates a working algorithm. There are rule-based algorithms and non-rule-based algorithms. The former are static rules set by humans. The latter are algorithms based on machine learning (a subfield of artificial intelligence).
The field of artificial intelligence involves the study and design of intelligent systems (intelligent agents) that perceive their environment and act to maximize their chances of success. It is also often described as a science that lets machines do things that would require intelligence if done by humans.
The term attrition rate comes from economics and describes the decline in the number of buyers of a product or service, for example, due to the age of the product or increased competition.
The term Big Data can also be understood as “large amount of data” or “mass data”. This data is either too large, too complex, too fast-moving or too weakly structured to be analyzed using manual data processing methods.
Black-Box Machine Learning
The term black-box machine learning refers to the part of machine learning where algorithms make decisions without the decision-making path being traceable, viewable, or transparent to the humans using the system.
B2B / Business-to-Business
A business-to-business relationship exists between two or more companies. This is contrasted with a business-to-consumer (B2C) relationship, which describes a business relationship between a company and an end customer.
Causality is the relationship between cause and effect. This is the sequence of events and states that relate to each other. Therefore, A is the cause of effect B and B is caused by A.
A correlation measures the strength of a statistical relationship between variables. In the case of a positive correlation, “the more variable A,the more B” or vice versa, and in the case of a negative correlation, “the more variable A,the less variable B”.
The churn rate indicates how many of a company’s customers have left over a certain period of time compared to the existing customer base.
Clustering is an umbrella term for various methods that belong to unsupervised learning and thus a common technique for statistical data analysis. In data science, it is used to gain valuable insights from our data. This method is used to find groups where the elements within a group are very similar (coherence), while the groups should differ from each other as much as possible (isolation).
Cross-selling describes the existing potential to sell another product or service to a customer who already buys another product or service.
Customer lifetime value is the present value of future cash flows attributed to the customer relationship. It is also described as the total value of a customer to a company over the entire duration of its business relationship.
In deep learning the algorithm receives raw data and decides for itself which features are relevant. It is a branch of machine learning that uses neural networks with many layers. Such a deep neural network analyzes data much like a human would look at a problem.
A decision tree is a diagram that can be used to determine a course of action that shows the statistical probabilities of a previously determined event. Because of its similarity to the plant that gives it its name, the diagram is called a decision tree, which is usually represented as a vertical diagram or with branches. Based on the decision itself, each “branch” of the decision tree represents a possible decision.
Dynamic pricing refers to a method of setting the price of a product or service based on the characteristics of the potential buyer or the prevailing circumstances (for example, demand and competitive situation).
Machine learning is the process by which computers change the way they perform tasks by learning from new data, without the need for a human to provide instructions in the form of a program. In this process, algorithms mimic the way humans learn and in this way gradually improve their accuracy.
The field of predictive analytics attempts to make predictions about the future from past data using various techniques such as data mining, modeling, machine learning, and others.
Predictive Sales Analytics
Predictive sales analytics is a subfield of predictive analytics that relates exclusively to forecasting in the area of sales.
Random Forest Method
The Random Forest method is an algorithm from the series of classification and regression methods that uses a combination of many different decision trees to make the best possible decision.
Supervised learning is a subcategory of machine learning. In supervised learning, there is usually an outcome variable that controls the learning process. In particular, it is characterized by using labeled data sets to train algorithms that can classify data or accurately predict outcomes. There are several supervised machine learning algorithms such as decision trees, K-Nearest Neighbor (KNN), Support Vector Machines (SVM), and Random Forests.
Unlike supervised learning, where there is a result variable that guides the learning process, unsupervised learning involves creating models of unlabeled data without predefined classes or examples. This means that no “supervisor” is available and learning must rely on heuristic guidance generated by the system by examining various sample data or the environment. These algorithms thus discover hidden patterns or data groupings without requiring any human intervention.