Loading 
Bonanza Offer FLAT 20% off & $20 sign up bonus Order Now
ITNPBD6
UK
University of Stirling
In modern day, performance may be seen as an evaluation of the stores that is indispensable and paramount , and which needs to be assessed and the progress of achieving the best organization objectives and goals [1].Performance determines the knowledgeability and feedback that tries to identify the most strengths and weakness of a certain organization performances and position that is always in a turbulent environment [2].Furthermore, performance management organizes individual and behaviors of the organization and try to improve the performance at a lower limit and upper limit of a given conglomerate [3].Scholars have suggested that this organization have efficient and effective performance to get higher profits ,better financial turnover and always a high productivity rate.
The key to success in any competition is to try and improve the performance of a given store in order to fully become knowledgeable ,having abilities and increase behavior of the staffs [4]. In this line ,the management at the top learns about how to face the current and future needs of their organization and in order to receive proper feedback and increase their profits for their business’s objectives and goals of the business motivates staffs to perform well ,increase competition score ,increase the performance of each store at a different location in different countries. This will help in implementing business strategy [5].
In order to perform well ,it is subjected to organization human resource or rather organization human power and their interaction with the resources, facilities and technology availability in the organization. The ability of the staffs to perform well is subjected to the skills and knowledge of the business understanding and application of the relevant skills [6]. Despite the resources of human power and motivation solely depends on the clearance space of the store that the staff occupies and the attitude towards the customers and set of policies and measurement to achieve goals and focusing on the individual performance of the store and a certain manager name. The important elements that always that always links the organization and the goals ,the determination of the performance goals and using the performance goals to evaluate the human resources in the organization. In [7], it recognizes the strengths and weakness of the organization. Moreover, performance evaluation that measures the efficacy of the organization. If the performance on one store improves ,the business gets higher revenue and if it goes on the opposite direction ,then the best losses a quite substantive amount of profits at any given time point. By enhancing the productivity of the staffs and all the resources and team building through acquiring of skills and increasing the commitment and the performance of the given firm. The ultimate goal is to try and evaluate the staff performance ,the store performing condition that will try to prevent undesirable losses [8].
Mining of data is a logical and continuous process that is used in searching through large amount of data in order to try and find useful data [9].The goal of this technique is to try and find patterns that originally were not there. The moment these patterns are discovered ,then the decision of the development is formed and deployed. The general steps for mining of the data are ;
Exploration is the first step of exploring of the data and its cleaned ,transformed into a more readable data and then the most important variable in the nature of the dataset that is always based in the problem to be solved in the identification of pattern [10]. The moment the data has been explored, redefined to form a specific variables ,the next step is to try and form a pattern which will make the best prediction. The last step is deployment of the desirable outcome.
The algorithms that are used for mining of the data are classification ,regression ,clustering ,neural network and the association rule [11]. In our case, we will consider the classification algorithms ,logistic regression and decision tree neural networks.
This is the most common type of classification that is applied in mining of the data. It employs a set of classification observations in order to develop a model that can correctly classify the population and records at large [12]. Examples of classification are credit risk application ,fraud detection and classification of the performance of the store. In this approach ,it mostly employs decision tree and neural network algorithms’ classification algorithms is based on learning and classifying. The learning of the training examples is analyzed by algorithm that specifically classify and then the test data are used in determining of the accuracy of the model. The algorithm encodes the parameters into a model that is called a classifier.
The technique that is adapted for prediction are called regression techniques. This analysis mostly uses models to map the relationship between one or many independent variables and depended variable. In mining of the data, the explanatory variable are the variables that are already present and known while a response variable is that variable that needs to predicted. The example of a prediction technique is predicting fuel prices and predicting of house prices.
This is a set of connected input or output that are each connected to a node and that they have weights present in them. In the learning these network adjusts the weights in order to enable the algorithm to predict correct class labels. These network are great in such that they drive their ability from human brains that consist of neurons. These algorithms work best for continuous valued inputs and outputs. A good example is image classification .The common type of neural network is backpropagation.
This research paper used secondary sources of data collection such as internet, websites, journals as well as articles about business performance from the store data [13]. This project used the store data which included variables such as location ,country, town, clearance space, car park and performance score. The study used nineteen variables namely: town, location country, store id ,demographic score, clearance space ,location,40minute drive, 30minute drive, 20minute drive, 10minute drive, car park, competition number ,competition score ,manager name and performance of the store.
The data used is efficient and more significant in trying to predict the performance of the store. The data types in our dataset consist of numerical and categorical. Our response variable is categorical in which it is two binary. The model will try to classify either bad or good .Other categorical variables are car park, location, manager name ,town and country’s remaining variables are integers .
In this section, the data exploration is done in depth. The visualization and descriptive statistics of the data will be done so that to grasp ideas of the data and how to interpret it when the modelling process begins.
The table below shows descriptive statistics of various numerical data in order to get their measure of central tendency and how they behave when the variable has extreme values.
From the summary of the data ,we can observe that the mean of the clearance space is 262.42 which means the expectation of having an average space is 262 units and the error is 51.72.The minimum units of the clearance space is 200 and the maximum units is 408.Only 75 percent of the stores have 292 units .The average number of staff present in the store is 13.35 and the variability of the number of staff is high this means that only 75 percent of the store staff are 8 people a figure that is actually lower with the maximum number of the staff being 600.The floor space of each store on average it has 15024.09 units with a variability of 2761.8.The minimum floor space is 10072 and the maximum number of floor space is 19869 units. The window in each space has average of 112 units and the error being 6.88.The minimum number of windows is 100 while the maximum number is 124 windows. The demographic store has on average 14.46 and the variability being lower of the number 2.87.The maximum number of demographic score is 19.The age of the store varies with a variability of 2.82 while on average the age of the stores is 6.53.The minimum number of age of the store is 2 years while the maximum number of the age of the store is 11 years. Only 75 percent of the stores have ages of 9 years.
After data summary ,we confirm through visualization the number of performance in relation to the competition score. The graph below shows the performance of the store against the clearance space.
The quality of the data should be of high value in order to generate appropriate results. This means that the preparation of the data needs to be done thoroughly in order to remove missing values, inconsistent and outliers in the data.
In this section, data preparation is the crucial stage of every analysis. This involves cleaning and wrangling of the data in order to get the transformed data in good condition. In our dataset there was no missing data this means that the data entry process was good [14] .Therefore, the next process was to identify the anomaly in the data and I found that on the variable staff the data was negative two which represent anomaly and inconsistent. This means we cannot have staff with negative number in which its not logic [15]. I went ahead and transform the data to be greater than zero in order to remove the inconsistent and the outlier. After removal of the inconsistence ,I plotted the histogram to confirm the transformation of the data. The figure below shows the histogram of the staff variable.
The section of data modelling basically focuses on selecting the best algorithm in order to model our data set. The models to be considered are logistic regression model, the decision tree and the neural network algorithms.
It is also known as the logistic model or the logit model. It analyzes the relationship between multiple independent variables and categorical dependent variable and try to estimate the probability of occurrence of an event by fitting data to a logistic curve [18]. The two types of logistic model are binary logistic regression and multinomial logistic regression. In binary logistic regression, the response variable is dichotomous while in multinomial logistic regression the response variable is two or more categories.
It represents a structured tree classifier that usually performs a split test in its internal node and try to predict a response variable in its leaf node [19]. Decision tree are simple hence transparent and are used widely in mining of data. In this report we employed a decision tree algorithms in classifying of whether the performance of the store will be good or bad. The proposed simplified decision tree that is most efficient in mining of the data that is based on entropy splitting criterion that is combined with pre pruning in order to limit the growth of the tree [20].
The artificial neural networks are likely more efficient in classification of algorithms .Neural network base on using back propagation in the classification of the data [21]. Once the data is fended in the network, the model compile and generate the output after carefully optimizing the learning rate [22]. There are many techniques that are used in optimization such as Adam,Relu and stochastic gradient descent. The graph below shows the structure and process of the neural network after compilation of the model.
In this report ,I analyzed data and try to classify whether the performance of the store will be good or bad. The data was first prepared and transformed into a well readable format [24]. The analysis was carried out using the three model that are logistic regression, the decision tree algorithm and the neural network. The logistic regression seems to work very well in the classification of the performance of the stores.
Are you stuck with an overly convoluted assignment based on a particular subject matter? Are you looking for a dedicated team of subject matter experts to help you through the hurdle? Take a look here. MyAssignmenthelp.co.uk is right here to back you up with the following services and beyond.
So, get in touch with us right away, place your order and have the best SME by your side to provide you with impeccable assignment help online.
Upload your Assignment and improve Your Grade
Boost Grades
