|
Topics:
|
|
|---|---|
Words: |
4016 |
Pages:
|
9
This essay sample was donated by a student to help the academic community. Papers provided by EduBirdie writers usually outdo students' samples.
|
Cite this essay
Download
Data Mining is implementing in wide variety of areas. There are many commercial data mining systems available in the world today. It is used in Loan payment prediction and customer credit policy analysis as far as financial data analysis is considered. In retail industry analysis of effectiveness of sales, campaigns is done with data mining. It plays an important role in providing the visualization tools in telecommunication data analysis. Biological data mining is a very important part of Bioinformatics. [10] In Biological data analysis, discovery of structural patterns and analysis of genetic networks and protein pathways is done with the help of data mining. It also contributes in intrusion detections. [10]
Bone diseases like trauma, inflammation, arthritis, osteoporosis, bone tumor, etc are nowadays very common in people. Not only these diseases but also the bone fractures are very much common in the people nowadays. Hence we came up with the system which considers and analyses the previously-stored symptoms of the bone disease patients and predicts the status of the new patient which determines the complexity, and type of the bone disease the patient is involved. As a result of our system, unnecessary tests on bones can be avoided.
Bone is one of the important components of a human body. It is the one which supports and protects various organs. And it also provides a structure for the body. Being so important to our body it may also get affected by various diseases or infections. So our primary objective is to identify those diseases or infections of humans at the initial state itself which will avoid the unnecessary tests on bones.
We in our project are considering the symptoms, signs, age found to be affected, and gender found to be affected, of the previously affected patients and when a new patient comes with the similar symptom or sign we consider the previously stored and analyzed data and predict whether the patient is affected by the similar disease or not. As a result of this, the person may avoid some of the unnecessary tests.
The report is well organized. It is divided into totally 9 chapters. The first chapter is the introduction which includes background, brief history of technology, applications, research motivation and problem statement, research objectives, contributions, organization of the report, and summary. The next chapter is Literature Survey, the other chapter is System Requirements Specifications, the other chapters are Design, Implementation, Testcases, Results, Conclusion, and References.
The research works have been found related to “Bone Disease Prediction Using Data Mining Techniques”. The dataset, the algorithms, the methodology used by the authors, and the observed results along with the future work is carried out in finding out efficient models of medical diagnosis for various bone diseases.
Bone diseases are mainly of the following types:
Bone disease is any of the diseases or injuries that affect human bones. In, Osteoporosis is a condition that weakens bones, making them fragile and more likely to break. More than 300,000 people receive hospital treatment for fragility fractures every year as a result of osteoporosis.
Here is a brief discussion about the work related to Bone Disease Prediction that has been already carried out in past few years.
Low Bone Mineral Density (BMD) is a risk factor for fracture and is considered as the important predictor of future fractures. The authors studied the relationship between Bone Mineral Density measurements at peripheral sites and subsequent fracture risk at the hip, wrist/forearm, spine, and rib in 149524 postmenopausal white women, without a prior diagnosis of osteoporosis. At enrollment, each participant completed a risk assessment questionnaire and had BMD testing at the heel, forearm, or finger. Main outcomes were new fractures of the hip, wrist/forearm, spine, or rib within the first 12 months after testing. The test is considered as the T-scores which is the measure of Bone Mineral Density for the prediction. The authors examined BMD measurement, Questionnaires, and Data analysis for the prediction purpose. The authors inspected the performance of the algorithms through evaluation criteria such as sensitivity and specificity
The authors studied that different markers of bone turnover predict the fracture in 1040 elderly women. The various markers considered were Serum bone-specific alkaline phosphatase and four different forms of serum osteocalcin (S-OC), and others as markers of bone resorption. They considered Sampling procedures, Bone markers formation, Bone markers resorption, Bone markers urine osteocalcin, and other measurements for the prediction.
The paper mainly focuses on low bone mineral density which is the important predictors of future fractures. The authors considered the association of Bone Mineral Density(BMD) with the risk of fracture of the spine or hip among a cohort of 2356 men and women aged 47–95 years, who were followed up by biennial health examinations. Follow-up averaged 4 years after baseline measurements of BMD that were taken with the use of Dual-energy X-ray Absorptiometry(DXA). Vertebral fracture was assessed using semiquantitative methods, and the diagnosis of hip fracture was based on medical records. Poisson and Cox regression analysis were the models used.
The authors trained an independent model based on a specific group of patients. They examined Comprehensive Disease Memory (CDM) which captures the characteristics for all patients to predict the disease. Bone disease memory (BDM) memorizes the characteristics of those considered individuals who suffer from bone diseases. Similarly, the Non-Disease Memory (NDM) memorizes attributes for non-diseased individuals. They have used Shallow Restricted Boltzmann Machine and 2-Layer Deep Belief Network for the prediction purpose.
This work applied many models to predict and prevent various bone diseases. In the respective research, the author has used oomph model as the opening to the proposed method, then single-layer and multi-layer learning approaches are introduced to construct the different disease memories. Finally, they have proposed their model that are focusing on the prediction and educational Risk Factor selection for bone diseases. The author analyzes the performance of the algorithms through evaluation criteria such as sensitivity to skewed class, sensitivity to noisy data, and parameter selection.
We consider Python interpreter for implementation purposes and Jupyter notebook web application for the better visualization of data and to check the accurate algorithm among Logistic Regression (LR), Linear Discriminant Analysis (LDA), K-Nearest Neighbor (KNN), Decision Tree (DT), Naïve Bayes (NB), Support Vector Machine (SVM) which we have used.
Jupyter Notebook: The Jupyter Notebook is an open-source web application that can be used to create and share documents that contain live code, equations, visualizations, and text.
The system predicts the type of bone diseases which are generally occurring in the human body by analyzing the dataset which can be divided into training and test set, based on the different number of features and by using more accurate algorithm. The prediction system predicts the two types of bone diseases such as Traumatic bone disease and Degenerative bone disease.
The Jupyter Notebook is an open-source web application which can be used to create and share the documents that contain live code, equations, visualizations, and text.
Fig 3.2.2.1. Jupyter web application
Fig 3.2.2.2. Python 3.6.4 Idle
The different modules are explained below:
In this section, we have discussed and organized the functional requirements for the prediction system. And also the System requirement for the prediction system which includes hardware and software requirements are specified in this section. These are the requirements that should be fulfilled to successfully complete this project.
Fig 4.1: Architectural Design
The above diagram Fig.4.1 demonstrates the architectural design for carrying out the prediction. The dataset (i.e Bone disease dataset) is fed to the ‘Training phase’ where six algorithms, kNN, Support Machine Vector, Decision Trees, Linear Discriminant, Logistic Regression, and Naïve Bayes are applied on the dataset to compare their efficiency. In the ‘Testing phase,’ the most efficient algorithm is selected among all of the six algorithms, then it is used to predict the type of bone disease when new data is fed for prediction.
Fig 4.2: Dataflow Diagram
The above diagram Fig.4.2. demonstrates the flow design for carrying out the prediction. Flow diagram is basically used to demonstrate how the data flows in the proposed system. The dataset is separated into the Training set and the Test set. The Training set is the set of data which is used to train a model. In training the model, specific features are picked out from the training set. Test data is the set of data on which the model is applied to predict the output. In the figure above, training set undergoes pre-processing to understand and select the special features and after pre-processing, the pre-processed data is fed to several algorithms, among which the most efficient is selected for the prediction. Test data is fed to the selected model to obtain the predicted output.
Fig 4.3: Class Hierarchy Diagram
The above diagram Fig.4.3. demonstrates the Class design for carrying out the prediction. Class diagram basically consists of all the classes defined in the project with its attributes and functions. In our project, there are four classes: main, get_data, predict_disease, and get_data1. get_data is the class which retrieves the data from the database. Predict_disease class is where prediction of bone disease is carried out.
Fig 4.4: Usecase Diagram
The above diagram Fig.4.4. demonstrates the use case design for carrying out the prediction. Usecase diagram is basically about how the user interacts with the system and the database. Here, the dataset is collected from several patients and these datasets are stored in the database. The database feeds these datasets to several algorithms. The predicted results are stored in the Excel sheet and the results can be viewed by the patient.
Fig 4.5: Sequence Diagram
The above diagram Fig.4.5. demonstrates the Sequence design for carrying out the prediction. The sequence diagram basically is used to demonstrate the sequence in which the whole process of disease prediction takes place. Here, data is collected from the patient and then stored in the database. After which the dataset stored in the database is fed to the algorithms and compared based on efficiency. The algorithm having best efficiency is then used to predict the result once the new data is fed to the selected algorithm. The predicted result and the dataset is stored in the database.
Fig 4.6: Activity diagram
The above diagram Fig.4.6. demonstrates the activity design for carrying out the prediction. The dataset is fed to six considered algorithms(kNN, SVM, LR, LD, DT & NB) and the best algorithm among all six is selected based on their efficiency. New data is fed as input to the selected algorithm and the result is obtained.
Dataset
The Dataset is collected from Sun Orthopedic Hospital, Mathikere. It has 47 rows, 29 attributes, and 2 classes.
Names do not contribute for the purpose of bone disease prediction. So this attribute is not used in our models. The dataset description is given in Table 5.1.1.
Table 5.1.1. Dataset Description
The algorithms which are used for the prediction of bone diseases are Decision Trees (DT), Logistic Regression (LR), Support Vector Machine (SVM), and K-Nearest Neighbor (K-NN). The description of these algorithms is given in the following section.
Logistic regression is a classification and predictive algorithm. LR is used to describe data and to explain the relationship between one dependent binary variable. There are one or more independent variables that determine an outcome. The binary logistic model is used to estimate the probability of a binary response based on one or more predictors. It is used to predict a binary outcome such as, “0” or “1” which may represent “Yes” or “No”, “True” or “false” in a given set of an independent variables.[9]
Logistic Regression Equation is shown in Equation (1) and its respective Sigmoid curve is shown in Fig. 4.2. (1)
Fig. 5.1.1.1. Logistic Regression Sigmoid Curve
LDA is most commonly used as dimensionality reduction technique in the pre-processing step for the pattern-classification and the machine learning applications. The goal is to project a dataset onto a lower-dimensional space with good class separability in order to avoid overfitting and also to reduce the computational costs.
The KNN algorithm is a non-parametric method used for classification and regression. The input consists of the k closest training examples in the feature space. The output depends on whether k-NN is used for classification or regression.[7] The K nearest neighbors are measured by a distance function, distance function considered is Euclidean distance.
The decision Trees algorithm can be used for solving regression and classification problems. Decision Tree creates the Training model which will be used to predict the class or value of target variables by learning decision rules inferred from the training data.
The Naive Bayes algorithm is a technique based on Bayes Theorem which is used for classification with an assumption of independence between class predictors. The Naive Bayes classification algorithm assumes that the presence of an exact feature in a class is dissimilar to the presence of any other features. The Naive Bayes algorithm model is easy to implement and useful for large datasets.
Bayes theorem describes the way of calculating posterior probability P(Cx|X) from P(C), P(X), and P(X|Cx) shown in equation (3.1).
Support Vector Machine is used for classification and regression analysis.
The implementation can be broadly divided into 6 parts:
1
2
1
Table 6.1. Testcases
Table.7.1. Accuracy and Error rate of Algorithms
The above table specifies the accuracy and error rate of the algorithms. The system checks for the accuracy of the algorithms. As a result, it stores the algorithm which is more accurate and having least error rate compared to others for the further prediction of newly entered Bone Disease data. In our case, LR algorithm is chosen as it having more accuracy and least error rate than others.
Fig.7.2. Comparative performance analysis of models
The above figure shows the comparative performance analysis of models. Accuracy and error rate are considered as the legends of the graph.
Fig.7.1. The Accuracy of all Algorithms
Fig.7.2. The Splitting of Training and Test Data
Fig.7.3. A Simple Login Application
Fig 7.4. Incomplete login
Fig.7.5. Enter Username and Password to login
Fig.7.6. Invalid Username and Password
Fig.7.7. Data entering interface
Fig.7.8. New Patient (Test Data) entry
Fig.7.9. Traumatic Bone Disease
Fig.7.10. New Patient (Test Data) entry
Fig.7.11. Degenerative Bone Disease
Fig.7.12. Description of Bone Disease
In this project we consider greater number of features and efficient algorithms to predict bone diseases more accurately. The system evaluates various data mining techniques such as Support Vector Machine, Logistic Regression, Decision Trees, and K-Nearest Neighbor. Through our evaluation, we found that Logistic Regression algorithm gives better accuracy than other data mining algorithms. Hence Logistic Regression algorithm is used for predicting the bone diseases. With the training dataset, we develop a model that predicts the type of bone diseases. The developed Logistic Regression model performs classification and prediction of the test dataset based on the training dataset. The prediction of bone diseases considered in our work are Degenerative and Traumatic diseases.
Get your paper done in as fast as 3 hours, 24/7.
Fair Use Policy
EduBirdie considers academic integrity to be the essential part of the learning process and does not support any violation of the academic standards. Should you have any questions regarding our Fair Use Policy or become aware of any violations, please do not hesitate to contact us via support@edubirdie.com.
Stuck on your essay?
