Exemple #1
0
        public void TrainClassifier(DataTable table)
        {
            dataSet2.Tables.Add(table);

            DataTable GaussianDistribution = dataSet2.Tables.Add("Gaussian");

            GaussianDistribution.Columns.Add(table.Columns[0].ColumnName);

            for (int i = 1; i < table.Columns.Count; i++)
            {
                GaussianDistribution.Columns.Add(table.Columns[i].ColumnName + "Mean");
                GaussianDistribution.Columns.Add(table.Columns[i].ColumnName + "Variance");
            }

            var results = (from myRow in table.AsEnumerable()
                           group myRow by myRow.Field <string>(table.Columns[0].ColumnName) into g
                           select new { Name = g.Key, Count = g.Count() }).ToList();

            for (int j = 0; j < results.Count; j++)
            {
                DataRow row = GaussianDistribution.Rows.Add();
                row[0] = results[j].Name;
                int a = 1;

                for (int i = 1; i < table.Columns.Count; i++)
                {
                    row[a] = VarianceAndMean.Mean(SelectRows(table, i, string.Format("{0} = '{1}'", table.Columns[0].ColumnName, results[j].Name)));

                    row[++a] = VarianceAndMean.Variance(SelectRows(table, i, string.Format("{0} = '{1}'", table.Columns[0].ColumnName, results[j].Name)));
                    a++;
                }
            }
        }
Exemple #2
0
        //الميثود الخاصة بعملية التصنيف
        public string Classify(double[] obj)
        {
            Dictionary <string, double> score = new Dictionary <string, double>();

            var results = (from myRow in dataSet.Tables[0].AsEnumerable()
                           group myRow by myRow.Field <string>(dataSet.Tables[0].Columns[0].ColumnName) into g
                           select new { Name = g.Key, Count = g.Count() }).ToList();

            for (int i = 0; i < results.Count; i++)
            {
                List <double> subScoreList = new List <double>();
                int           a = 1, b = 1;
                for (int k = 1; k < dataSet.Tables["Gaussian"].Columns.Count; k = k + 2)
                {
                    double mean     = Convert.ToDouble(dataSet.Tables["Gaussian"].Rows[i][a]);
                    double variance = Convert.ToDouble(dataSet.Tables["Gaussian"].Rows[i][++a]);
                    double result   = VarianceAndMean.NormalDist(obj[b - 1], mean, VarianceAndMean.SquareRoot(variance));
                    subScoreList.Add(result);
                    a++; b++;
                }

                double finalScore = 0;
                for (int z = 0; z < subScoreList.Count; z++)
                {
                    if (finalScore == 0)
                    {
                        finalScore = subScoreList[z];
                        continue;
                    }

                    finalScore = finalScore * subScoreList[z];
                }

                score.Add(results[i].Name, finalScore * 0.5);
            }

            //اختبار اكبر قيمة احتمال للاصابة بالمرض
            double maxOne = score.Max(c => c.Value);
            double minOne = score.Min(c => c.Value);
            var    name   = (from c in score
                             where c.Value == maxOne
                             select c.Key).First();



            //خرج الخوارزمية وهي قيمة الحالة وقيمة الاحتمال
            string r = "";

            if (name == "1")
            {
                r = "Positive";
            }
            else
            {
                r = "Negative";
            }

            return("The Result is " + r + " and the probability is " + maxOne);
        }
Exemple #3
0
        //الميثود الخاص بعملية تدريب الخوارزمية على البيانات
        public void TrainClassifier(DataTable table2)
        {
            dataSet.Tables.Add(table2);

            //table
            //استخدام التوزع الغوصي
            DataTable GaussianDistribution = dataSet.Tables.Add("Gaussian");

            GaussianDistribution.Columns.Add(table2.Columns[0].ColumnName);

            //columns
            for (int i = 1; i < table2.Columns.Count; i++)
            {
                GaussianDistribution.Columns.Add(table2.Columns[i].ColumnName + "Mean");
                GaussianDistribution.Columns.Add(table2.Columns[i].ColumnName + "Variance");
            }

            //calc data
            var results = (from myRow in table2.AsEnumerable()
                           group myRow by myRow.Field <string>(table2.Columns[0].ColumnName) into g
                           select new { Name = g.Key, Count = g.Count() }).ToList();

            for (int j = 0; j < results.Count; j++)
            {
                DataRow row = GaussianDistribution.Rows.Add();
                row[0] = results[j].Name;
                int a = 1;

                for (int i = 1; i < table2.Columns.Count; i++)
                {
                    //حساب التبياين
                    row[a] = VarianceAndMean.Mean(SelectRows(table2, i, string.Format("{0} = '{1}'", table2.Columns[0].ColumnName, results[j].Name)));
                    //حساب المتوسط الحسابي
                    row[++a] = VarianceAndMean.Variance(SelectRows(table2, i, string.Format("{0} = '{1}'", table2.Columns[0].ColumnName, results[j].Name)));
                    a++;
                }
            }
        }