Targeted marketing strategy is a prominent topic that has received substantial attention from both industries and academia. Market segmentation is a widely used approach in investigating the heterogeneity of customer buying behavior and profitability. It is important to note that conventional market segmentation models in the retail industry are predominantly descriptive methods, lack sufficient market insights, and often fail to identify sufficiently small segments. This study also takes advantage of the dynamics involved in the Hadoop distributed file system for its ability to process vast dataset. Three different market segmentation experiments using modified best fit regression, i.e., Expectation-Maximization (EM) and K-Means++ clustering algorithms were conducted and subsequently assessed using cluster quality assessment. The results of this research are twofold: i) The insight on customer purchase behavior revealed for each Customer Lifetime Value (CLTV) segment; ii) performance of the clustering algorithm for producing accurate market segments. The analysis indicated that the average lifetime of the customer was only two years, and the churn rate was 52%. Consequently, a marketing strategy was devised based on these results and implemented on the departmental store sales. It was revealed in the marketing record that the sales growth rate up increased from 5% to 9%.

Market Intelligence is knowledge extracted from numerous data sources, both internal and external, to provide a holistic view of the market and to support decision-making. Association Rules Mining provides powerful data mining techniques for identifying associations and co-occurrences in large databases. Market Basket Analysis (MBA) uses ARM to gain insights from heterogeneous consumer shopping patterns and examines the effects of marketing initiatives. As Artificial Intelligence (AI) more and more finds its way to marketing, it entails fundamental changes in the skills-set required by marketers. For MBA, AI provides important ways to improve both the outcomes of the market basket analysis and the performance of the analysis process. In this study we demonstrate the effects of AI on MBA by our proposed new MBA model where results of computational intelligence are used in data preprocessing, in market segmentation and in finding market trends. We show with point-of-sale (POS) data of a small, local retailer that our proposed “Åbo algorithm” MBA model increases mining performance/intelligence and extract important marketing insights to assess both demand dynamics and product popularity trends. Additionally, the results show how, as related to the 80/20 percent rule, 78% of revenue is derived 16% of the product assortment.

La presente tesis de investigación, tiene como objetivo proponer un método para la segmentación de clientes, incorporando la predicción del valor monetario del cliente como una variable de segmentación, para tal fin, se propone una metodología cuantitativa, en la que los datos a utilizar corresponden a las transacciones de una tienda en línea de regalos para toda ocasión de Reino Unido, denominada "Online Retail II", que consta de un total de 5.833 clientes y 1.067.371 registros; a partir de los cuales se realiza un proceso de caracterización de los datos, seguido de la predicción del valor monetario de cada cliente utilizando técnicas estadísticas y de aprendizaje de máquinas, que posteriormente se incluye como variable en el proceso de segmentación.Finalmente, se hace un comparativo entre los resultados de segmentar clientes sin incorporar la predicción del valor monetario y la segmentación de clientes incorporando la predicción del valor monetario; con lo que se concluye que el método propuesto, utilizando el algoritmo de Vecinos más cercanos para la predicción del valor monetario del cliente, al incorporarlo en la segmentación de clientes, logra un desempeño económico entre 10% y 20% mejor que segmentar sin incorporar esta variable.

Finding outliers, rare events from a collection of patterns, has become an emerging issue in the area of machine learning concerned with detecting and eventually removing anomalous objects in data. A key challenge with outliers/anomalies detection is because they are not a wellformulated issue. Outliers are defined as the extreme values that deviate from the overall patterns in data; they may indicate experimental errors, variability in measurement, or a novelty. Detecting outliers in large databases can lead to the discovery of hidden knowledge. However, identifying and removing outliers often helps to ensure that the observations represent the problem correctly. Though there are several techniques for detecting outliers/anomalies in a given database, thus, no single technique is proven to be the standard universal choice. Depending on the nature of the target application, different implementations require the use of different outlier detection methods. The clustering method is a very powerful method in the field of machine learning and defines outliers in terms of their distance to the cluster centers. In this study, we propose a clustering-based approach to identifying outliers in a retail point-of-sales dataset. To select the best clustering algorithm for the purpose, two algorithms are applied, K-means for hard, crisp clustering, and (FCM) Fuzzy C-means for soft clustering. The experimental results show that the K-means algorithm outperforms the (FCM) Fuzzy C-means algorithm in terms of outlier detection efficiency, and it is an effective outlier detection solution.