Applications of Data Mining to Reward points collection service

Continuation of use cases:   

Machine learning techniques form an altogether separate category of their own. The traditional data mining methods used clustering and statistics which are relevant to machine learning, but we did not include the neural networks with data mining, and we call it out with others in this category. Machine learning is very helpful to inform users about their activities that generate the most appreciation and the changing of these activities depending on the audience. It can also detect fraud in the employee appreciations which may be of interest to employers. For example, Feedzai uses real-time behavioral profiling as well as historical profiling that has been proven to detect 61% more fraud than earlier. Discovering groups, searching and ranking are a few more examples.

Regions of interest are used to determine space and time focus on appreciation activity. This is helpful to detect events that would have otherwise gone unnoticed as a flurry of activities on the reward points table. Together with the classifier and this regressor, the latent events can be detected thus eliminating the need to hold formal events and determine winners.

One of the aspects of using neural networks with employee appreciation data is that the management can gain insights that would not otherwise have been possible with formal interactions. By classifying reward points based on vector features and using softmax classification, the neural networks can detect the hidden appreciation. Each neuron assigns a weight usually based on probability for each feature and the weights are normalized across resulting in a weighted matrix that articulates the underlying model in the training dataset. Then it can be used with a test data set to predict the outcome probability. Neurons are organized in layers and each layer is independent of the other and can be stacked so they take the output of one as the input to the other This is a technique that has found applications in a variety of domains starting from natural language processing.

Neural networks can be applied in layers and they can be combined with regressors so the technique can be used for a variety of use cases. There are four different types of neural networks. The fully connected layer connects every neuron in one layer to every neuron in another layer. This is great for rigorous encoding, but it becomes expensive for large inputs and scalability.  The convolutional layer is mostly used as a filter that brings out salient features from the input set. The filter sometimes called kernel is represented by a set of n-dimensional weights and describes the probabilities that a given pattern of input values represents a feature. A deconvolutional layer comes from a transposed convolutional process where the data is enhanced to increase resolution or to transform. A recurrent layer includes a looping capability such that its input consists of both the data to analyze as well as the output from a previous calculation performed by that layer. This is helpful to maintain state across iterations and for transforming one sequence to another.

The choice to apply machine learning techniques is dependent both on the applicability of the algorithm as well as the data.