Browsing by Author "Nurmambetov D."

Now showing 1 - 3 of 3
  • Thumbnail Image
    ItemOpen Access
    KAZAKH NAMES GENERATOR USING DEEP LEARNING
    (ВЕСТНИК КАЗАХСТАНСКО-БРИТАНСКОГО ТЕХНИЧЕСКОГО УНИВЕРСИТЕТА, №4, 2020) Nurmambetov D.; Dauylov S.; Bogdanchikov A.
    In recent years, sentiment analysis of e-mail messages or social media posts is becoming very popular. It can help people define if they are reading something positive or negative. On the same time, there are some services on the Internet that can help you find or create a new name. When processing the creation, they check the name in other popular languages, so your name does not mean inappropriate things in other languages. For this they bill for 25 thousand US dollars. If there are such services, then there is a demand. In this study, sentiment analysis of e-mails was implemented with using StanfordNLP [1] lemmatizer and classic machine learning algorithms as a classifier. It is applied to real e-mails from Russian speaking mailbox, which means there are both English and Russian messages. Thus, language identification is also added as preprocessing step. In this study only binary sentiment analysis was made, but it can be improved with adding several emotions to be detected. Then another model generates Kazakh names using neural networks, where all Kazakh names data has been collected through various websites. The sentiment analysis model gives 81% accuracy and the joint use of two models allow us to generate new Kazakh names, which are checked with Russian language if they mean something inappropriate. The result can be improved with checking with other languages.
  • Thumbnail Image
    ItemOpen Access
    Planning and data analysis system using machine learning algorithms
    (2020) Nurmambetov D.
    The information system, which includes data analysis, data clustering, as well as sales amount prediction using machine learning algorithms will be described in this study. The system is intended for work in a manufacturing company that sells certain products to retail outlets. According to the Forbes article, “Ten Ways Big Data Is Revolutionizing Supply Chain Management” [1], companies tend to integrate big data and advanced analytics into optimization tools. Thus, the idea of the system was proposed to increase efficiency of business processes. The system includes a method which can be used for outlets clustering using unsupervised and supervised machine learning algorithms comprising 2 steps — data partition using unsupervised Gaussian Mixture (GM) Model clustering algorithm based on outlets sales amount and further partition for one of them using Logistic Regression (LR) and Neural Networks (NN) classification algorithms, which predict whether outlets will achieve monthly sales plan. The proposed algorithm was tested on real sales data and formed 3 clusters according to business needs. Sales plan on gave up to 74% accuracy. Also, system should include achievement prediction plans settling algorithm based on sales forecasting per each outlet. This part was implemented using regression ensembling algorithm XGBoost, time series model SARIMA and recurrent neural network LSTM model. Sales forecasting was based on weekly sales data for 5 years in the outlet and predicts future sales amount values. Models were compared in terms of Mean Absolute Percentage Error (MAPE) and running time. From the results, it was decided to choose LSTM model which gave 17% MAPE and good running time to embed to the system with possible improvements.
  • Thumbnail Image
    ItemOpen Access
    SEGMENTATION OF OUTLETS USING MACHINE LEARNING CLUSTERING AND CLASSIFICATION ALGORITHMS
    (СДУ хабаршысы - 2019, 2019) Nurmambetov D. ; Bogdanchikov A.
    Abstract. Segmentation of retail outlets in terms of manufacturing companies’ strategy applied in sales amount and trade activities for each of them is very important. Directed investments into outlets help companies to make more profit and decrease expenses. This study presents a method which can be used for outlets clustering using unsupervised and supervised machine learning algorithms comprising 2 steps – data partition using unsupervised Gaussian Mixture (GM) Model clustering algorithm based on outlets sales amount and further partition for one of them using Logistic Regression (LR) and Neural Networks (NN) classification algorithms, which predict whether outlets will achieve monthly sales plan. Previously, clustering was made without any special methods and clusters were formed using some agreed threshold values for outlets sales amount. The proposed algorithm was tested on real sales data and formed 3 clusters according to business needs. Sales plan achievement prediction gave up to 74% accuracy.