Subtitle as needed

Subtitle as needed (paper subtitle) AbstractChronic Kidney Disease (CKD, also known as Chronic renal disease) is a communal problem to public with an escalating in either technologically advanced or advancing countries. 10 of the inhabitants worldwide is affected by chronic kidney disease (CKD), and masses die each day because they do not have access to reasonable treatment or due to lack of awareness. KeywordsChronic Kidney Disease Blood Potassium Levels Diet plans Machine learning Potassium zone Introduction Chronic kidney disease (CKD) is defined as kidney structural damage caused by disintegration function of kidneys and is usually measured with GFR (Glomerular Filtration Rate), where the Glomerular Filtration Rate drops, (GFR) 60 ml/min/1.73 m2 for three months or so. Such degeneration is problematic to waste and excess fluids formation in the body and impacts the performance of the body, potentially prominent to complications. It often remains overlooked and undiagnosed until the condition of the individual gets worse slowly over time. The disease can reach to end-stage renal disease (complete kidney failure) whichtakes place when kidney function gets deteriorated to a point where dialysis or kidney transplantation turns out to be the only way for survival. People suffering from blood pressure, diabetes and having a family background of being affected by such diseases or suffering from chronic kidney disease earlier are most likely to suffer from kidney diseases. The rate of death occurrence rose to 956,000 in 2013 from 409,000 in 1990 1. It is considered to be a life-threatening disorder affecting a large number of people. However, this rise in data volume automatically involves the data to be repossessed when needed. Health administrations today are capable of generating and accumulating a bulk amount of data.With the help of data mining and classification techniques in medicinal applications, it is possible to identify relationships and models that support forecasting and decision-making process for analyzing and action planning 2, an action planning may include the do and dont(s) of daily life activities for instances suggesting a meal or diet of an individual. For a CKD patient, following an appropriatediet plan can help to lessen the growth of CKD. So, it is very necessary to identify a suitable diet based on the patients health condition such as based on the estimated Glomerular Filtration Rate (eGFR), which can be categorized into five stages such as stage0, stage 1, stage 2, stage 3 and stage 4. Till stage 2, patients are considered to be within the safe range or they are able to cope up with the renal functions without gathering excretory products savor potassium or surplus urea in the blood. Henceforth, patients in the stage 0, stage 1, and stage 2 dont require any vital changes in their diet plan. But for patients in the stage 3 and 4 are in difficulty of keeping up the balance of minerals, electrolytes, and liquids inside their body. Be that as it may, for patients in the stage 3 and 4 are in trouble of keeping up the adjust of minerals, electrolytes, and fluids inside their body. Age-Adjusted Prevalence of CKD Stages 1-4 by Gender 1999-2015 3 Diet plans of CKD patients not simply rely upon the stage of the disorder yet additionally with different conditions, such as the level of blood potassium, urea, sodium etcetera 4. In this paper, diverse data mining techniques have been executed on a dataset containing data about patients determination for CKD. These methods are Nave Bayes, Support Vector Machine, and Multilayer Perceptron, but the core center is around blood potassium level to distinguish the reasonable eating routine arrangement for a CKD patient. The paper is divided into five sections. Section II represents the relevant works done so far for prediction of CKD using data mining techniques and how it is implemented in dietary system. In Section III, methodology consists of system model of the predictive diet plan and a brief description on how the algorithm are implemented on the model, this is done to provide a better understanding of the whole paper. Experimental results and result analysis are presented in the Section IV. Section V, consists of conclusion and additional future scope. Related Work In recent days, data mining techniques is contributing a lot in the health care system to solve or detect various disease. A number of researchers have used different algorithms and methods to classify or to detect CKD, Diabetes and other diseases as well. In 5, proposed a work using Nave Bayes with OneR attribute selector for detecting CKD. It helps to prevent renal disease reaching complicated stage. The recommended system extracts rules for the current stage so that required treatment is taken accordingly. In 6, the authors implemented an automated machine learning model to predict CKD and discover 24 attributes associated to it. Feature selection was performed to choose the important attributes for detection and then based on their predictability rank them. Different predictive attributes were identified. The solutions were then evaluated using three different classification algorithms such as k-nearest neighbor, random forest and neural networks. In 7, designed a food recommendation service which prepares a custom-made service for patients with coronary heart disease. It prepares the diet by considering some important information such as vital sign, specific food choice, and previous family history of the patient. The service provided is personalized unlike the conventional service that are mostly used. In 8, classification techniques are built by using Wrapper method. The method reduces the number of attributes to predict CKD. Wrappersubset attribute evaluator and bestfirst search are used and it improved the result of detecting CKD on reduced dataset. In 9, the researchers used Artificial Neural Network (ANN) algorithm and Support Vector Machine (SVM) algorithm to predict renal related disease. The main aim of the research was to relate the performance of the two algorithms based on their execution time and accuracy. It is observed that ANN algorithm performs better than SVM algorithm with an accuracy of 87. In 10, the authors presented a diagnosis method over fuzzy logic. MATLAB software was used which has in built fuzzy logic toolbox to check the condition of the kidneys of a patient. Data used for the research is collected from kidney patients of Birdem Hospital, Dhaka through several diagnosis reports. Seven attributes have been considered and the healthiness is calculated in a boundary from 0 to 10 for the analysis. In 11, the authors used Self-Organizing Map (SOM) and K-mean clustering for food clustering analysis and implemented Food Recommendation System (FRS) for diabetic patients. It recommends the substituted foods based on the connection of eight significant nutrients of a diabetic patient. Nutritionists evaluated the FRS which performed well and it is useful for diabetic patients. In 12, the researchers developed a machine learning model to increase the quality of CKD diagnosis by using feature selection and ensemble learning. In this paper, to improve the classification of CKD, Correlation-based Feature Selection (CFS) was used for features selection and AdaBoost was used for ensemble learning. Classifiers such as KNearest Neighbor algorithm (kNN), Naive Bayes and Support Vector Machine (SVM) were used. The best result was obtained by kNN classifier with CFS and AdaBoost with 0.981 accuracy rate. Methodology This study uses two publicly accessible datasets 13 14 which is taken from the UCI repository and Data World website, which consists the records of 400 patients collected from Apollo Hospital India and 61 varieties of raw food. Data Preparation for Patients Dataset Originally. the dataset of patients contains 25 attributes with people aged between 2 to 90 years old. Attributes and values are show in the Table 1. In the dataset, there are – Instances 400 (250 ckd, 150notckd) Attributes 24 Zone class 25 (11 numeric, 14 nominal) TABLE I. Attributes Description AttributesDescriptionBlood pressure (mm/Hg)Numerical ValuesSugarNominal Values (0,1,2,3,4)AlbuminNominal Values (0,1,2,3,4)red blood cellNominal Values(normal, abnormal)blood glucose random (mgs/dl)Numerical Valuesblood urea (mgs/dl)Numerical ValuesSerum creatinine (mgs/dl)Numerical ValuesPotassium (mEq/L)Numerical ValuesHemoglobin (gms)Numerical Valuespacket cell valueNumerical Valueswhite blood cell count (cell/cmm)Numerical ValuesHypertensionNominal Value (Yes, No)diabetes mellitusNominal Value (Yes, No)appetiteNominal Value (Good, Poor)pedal edemaNominal Value (Yes, No)ClassNominal Value (ckd,notckd) MissingValues Exist. Class Division ckd and notckd. After the evaluation of CKD and NOTCKD has been done we included an extra column named as ZoneClass. Based on the potassium level in blood, the instances have been categorized into four categories such as – If the blood potassium level is between 0 to 3.4, then it is identified as LOW. If the blood potassium level is between 3.5 to 5, then it is identified as SAFE. If the blood potassium level is between 5.1 to 6, then it is identified as CAUTION. If the blood potassium level is greater than or equals to 6, then it is identified as DANGER 15. System Architecture of Recommended Food System For this study, the range above and equal to 5.1 is categorized into 1 level named as ALARMING. To add the new column ZoneClass, we used an add-in of Microsoft Excel named Power Query. The formula used to evaluate the column is given in the fig. 2. We used Weka tool, therefore the missing values were not cleaned or randomized. Query to categorize the levels Data Preparation for Raw Food Dataset Initially, the dataset consists of 16 attributes including the name of the food, total, fats, sodium level, calories, potassium level and so on. In the final dataset, an extra attribute of blood potassium level has been added as the core targeting attribute, based on which suitable foods are recommend to CKD patient. Therefore, we are classifying it in 3 levels as LOW, SAFE and ALARMING. Where, if potassium in food, 5.1, then it is suitable food for LOW. 3.5 and 5.0, then it is suitable food for SAFE 0 and 3.5, then it is suitable food for ALARMING. To add the new column in the original dataset, we used add-in of Microsoft Excel named Power Query and generated a query shown in fig. 3. Query to categorize the levels Classification Methods Classification procedures that is also known assupervised learning techniques are most usually used in data mining to classify the data in a raw data. Some classification methods used are Bayesian classification, decision tree, support vector machine (SVM), neural networks, association-based classification. In the study, various classification methods are applied on the modified data. Three classifiers have been selected that were based on the attributes of dataset. They are- Nave Bayes, Random Forest and Support Vector Machine. I. Nave Bayes The Naive Bayesian classifier 15, is based on Bayes theorem with independence assumptions between predictors. A Naive Bayesian model is easy to build, with no complicated iterative parameter estimation which makes it particularly useful for very large datasets. Despite its simplicity, the Naive Bayesian classifier often does surprisingly well and is widely used because it often outperforms more sophisticated classification methods. P(xc) P(c) category. The system will predict the diseases the user is most likely to get, and will also recommend a P(cx) P(x) Here, P(cx) is the posterior probability of class (target) given predictor (attribute) P(c) is the prior probability of class. P(xc) is the likelihood which is the probability of predicator given class. P(x) is the prior probability of predicator II. Random Forest Classifier Random Forest Classifier is ensemble algorithm. Ensemble algorithmsare those which combines more than one algorithms of same or different kind for classifying data. Random forest classifier creates a set of decision trees from randomly selected subset of training set. It then aggregates the votes from different decision trees to decide the final class of the test data. III. Support Vector Machine Support vector machine is a machine learning tool based on statistical learning theory. It uses a nonlinear mapping to rehabilitate the data in to an upper dimension. It is based on the conception of decision planes that identifies decision boundaries. Decision plane is a discrete hyperplane – created in the descriptor space training data and compound, are classified based on the side of hyper plane they are located. The SVM takes data as input and predicts for each of them, which of two possible classes comprises the input and for which SVM is a non-probabilistic binary linear classifier. Training time for the SVM is extremely slow but they are extremely accurate in making predictions. Recommended Food In this step, both the preprocessed datasets are merged together in the power query, ultimately recommending food for the CKD patients. The food is recommended based on the three classified levels. The recommendation is done on the basis of, high potassium food for the ones with low potassium level, safe for the ones with safe level and low for the ones with alarming blood potassium level. Experimental Results and Evaluation In this study, 10-fold cross validation technique is used on the model, where 9 folds are used for training and 1-fold for testing, the overall process is repeated until all the 10 individual folds have been used for testing and the results are evaluated using Nave Bayes algorithm gives an accuracy of 95.5, Support Vector Machine(SVM) gives an accuracy of 98.25 and Random Forest algorithm gives an accuracy of 99.75. It is observed the Random Forest algorithm performs better to predict CKD than Nave Bayes and SVM. Fig. 5 shows comparison between the algorithms in terms of their accuracy. Accuracy in percentage for SVM, Nave Bayes and Random Forest. After achieving the number of CKD patients, we included the potassium intake column where the blood potassium level is taken as the main contributing attribute. The fig. 6 shows CKD patient in the 3 classified levels. 250 patients are divided into 3 levels based on blood potassium As our main objective to recommend food for the CKD patients based on their level of potassium in blood. After implementing the merging query on the Patients dataset and Raw food dataset, 61 food were recommended for the above classified levels. Table 2, shows the results or number of food items recommended in different level. TABLE II. Recommended Food LowSafeAlarmingApple Asparagus Banana Bell pepper Blue Crab Broccoli Cantaloupe Carrot Catfish Cauliflower Celery Clams Cod Flounder/sole Grapes Halibut Green Beans Haddock Honey dew melon Kiwi fruit Lobster Mushrooms nectarine Ocean preaches Orange Orange roughly Oyster Peach Plums Pollock Potato Rainbow trout Rockfish Salmon Atlantic Salmon Pink Scallops Shrimp Summer squash Sweet potato Sweet corn Sweet cherries Tomato Tuna WatermelonAvocado Cucumber Grape fruit Green cabbage Leaf Lettuce Onion Pear Radishes StrawberriesGreen Onion Iceburg lettuce Lemon lime Pineapple Swordfish Tilapia Recommended food into 3 classified levels. Conclusion Future Work In this study, we proposed a predictive approach using machine learning algorithm to identify CKD and NOTCKD patients, where Random Forest was more accurate than SVM and Nave Bayes, with an accuracy of 99.75. Based on the obtained results we recommended food for different level of CKD patients using blood potassium level. As our dataset was limited to 61 foods, therefore for future work this approach can be adapted to recommend food on larger dataset as well. References GBD 2013 Mortality and Causes of Death Collaborators, Global, regional, and national agesex specific all-cause and cause-specific mortality for 240 causes of death, 19902013 a systematic analysis for the Global Burden of Disease Study 2013, Lancet, vol. 385 (9963), p. 117171. J.C. Prather, D. F. Lobach, L. K. Goodwin, J. W. Hales, M. L. Hage, W. E. Hammond, Medical data mining knowledge discovery in a clinical data warehouse, Proc AMIA Annual Fall Symposium, pp. 101105, 1997. Picture available on – https// Ministry of Health, Nutrition Indigenous Medicine, Sri Lanka, Dietary Guidelines Nutrition Therapy For Specific Diseases, Online. Available HYPERLINK http// Uma N. Dulhare and Mohammad Ayesha, Extraction of Action Rules for Chronic Kidney Disease using Nave Bayes Classifier, International Conference on Computational Intelligence and Computing Research (ICCIC), A. Salekin and J. Stankovic, Detection of Chronic Kidney Disease and Selecting Important Predictive Attributes, pp. 262270, IEEE, Oct. 2016. J. H. Kim, J. H. Lee, J. S. Park, Y. H. Lee, and K. W. Rim, Design of Diet Recommendation System for Healthcare Service Based on User Information, Proc. 4th Intl Conf Computer Sciences and Convergence Information Technology, pp. 516-518, November 2009. N. Chetty, K. S. Vaisla, and S. D. Sudarsan, Role of attributes selection in classification of Chronic Kidney Disease patients, in Computing, Communication and Security (ICCCS), 2015 International Conference on, pp. 16, IEEE, 2015 Dr. S. Vijayarani and Mr. S. Dhayanand, KIDNEY DISEASE PREDICTION USING SVM AND ANNALGORITHMS,International Journal of Computing and Business Research (IJCBR), vol. 6, no. 2, 2015. S. Ahmed, M. T. Kabir and N. T. Mahmood, Diagnosis of kidney disease using fuzzy expert system., in 2014 8thInternationalConference on Software, Knowledge, Information Management and Applications (SKIMA), Dhaka, 2014, December. M. Phanich, P. Pholkul, and S. Phimoltares, Food Recommendation System Using Clustering Analysis for Diabetic Patients, 2010 Int. Conf. Inf. Sci. Appl., pp. 18, 2010. Potassium and Your CKD Diet, The National Kidney Foundation. Online. Available https// Accessed 24- Aug – 2017. http// Accessed 20th March, 2017 RTyLWUPvvVbbb )F)55U Whxcb0
7)_ UyE8gA4lfWWJVUQkaUQbRZGGUrNZ9frv(5xBkz QJOJ5TJ_KWvwkzroniS1PvWT4nvzLcbshz5B(99Y_,U(mZk) y OQM-ar)ip u4S7ui(-lxy06MEEU 2JWgGbcZJ_5qD3 4o86,wTVRU_pa,YM6cQQJ4W9gM TUDLZIaKkyaNAIvj51.VZdV)TllW
MmMJcUuZv)VZtPu Cy8Y0.Wf mU,2c S_LK6Ga/xkCijZD)s74SDsGUSnXCH5mf/TAyUOUr-VXGS1gVN6Pops6kr)IG,2cV4.NyUX1OTWDA2vi)mxc.PSnOK.AUUgXe([email protected] KWin( X(Quv3uJReePs/RmQIG(,Q)-)cvEYgr0bWGW1koXw).B YY3AMU/mYy9Upzrwjk3OF
avQneCE7BMiQZqP3G3GS)[email protected],V(F-ZyXrgNGu5mbWvhjNMuMdDsQAn A0QqWbmKRYK
Nm08K)YKj7r)7nmF-iSRen)VY35lbm1k [email protected]
jBmqo6xU26ST3tO zynKnyZm1qTTrSwcvP-JJSogq2T/e/2O4zzt(2Ow w5XY9JnYoabZrLwtz.XWK.PJ)i2unz)V
h_js [email protected] P2wmLIS5wlmzqjM/M9WyY2Mofo,TZ5WcJ6HEO3uyrpBMRE34gnq)[email protected],UTHE1QNR/gszTeNTlVL_ptL5mdW1GnG_eGCJ/Uac([email protected](PEe4AeZR3w RlVyh6FV2G9lMeN)UwKwOM9-F2E)E JJnP Cn(DhVWDnc O o()-HpNZC6Ao RE CGVsEh C- /V4a2yM-ly(A7 [email protected] oE5ad5lf2s_0 CQmBX 4/o [email protected],y fdWydwXfn
wpS [email protected] .IR01NM9zLsqbudnOG nUot/fxu2H8lNuRKYXfYjY5Hvrv6D(yQtMohDSW33OrObJDOvxiMeunG6mK5nk0WYqk7n8M_GpyNUkcl QREbRKRu WO6iiJEHnZVbblMm/Dwez-W e5ov(d6im(S7irPWXSzTivLaJyt/C TVCkWv_lp(Dto rp mNk kor_L1lyM6 Q9F S)[email protected]@@5IZMv17wKQ o(wVzbNiHgVDqK oUKiH/_P_I.qSw-QKns GyPied/m7Qtl3zjT1K5wHuSK5- [email protected]@5qILKACEneEhQSynEnOXs i3w HHU i7w.w wmlmkikgKPng T7GwQvwvNmUHAaytZ @9fnrnGZvrEFJh
yHi4mHQoyRJVu/z_bOqMtn/q-cNiTnmyT1/wPKvH9S3(_yNENMnS/XM_6-qM_M6h5.w UkiEwKfsLJq K/ gEMoKkf1h bnMovK 2bwCkw AGsKiqqH/nEen4KugQj2f_/bGwQsn.nYHn–EBioXh5.w [email protected](JGtGwQvqwVHAy6w @99ssiWAw [email protected]
GwqnW yC @9a oG7 yAw A0h7t nAqsyA-.w o4y yC @9a oG7 yAw A0h7tNWfgkS A0h7tw6 yC77/O 8qvMT OOkCs i5qm-YENtE

jGHj7P7jnkgkK3iyHakez.vabmAw8x2KlMIvxKJi.yvnjP MTvVigwXJXBdmTEnmj 4(7jY_kXmcLNKz6s5nafEAm(IAJ0coyKJQL4WDM6/WZ1Z54n zXVIWSN7oGUx4ne Tf2o1dgEbCloiYmmdOyizmjyJiP vT t,y9oPy/o gEov,Trm6hGo,VTWs_GEneLl2Ee.,s7bz [email protected] iyHBbI-17kYJNIi-Iny,wgk/y(mkKHfuMmemx1 VnGkxlKKik4bMek_yFsyw4._Ylv6YiGk9 nAk/6QgkKv-QgC nA-svvLWsow0g-_RYMChz-/c6eqMUZyDTyC PEme GkJmums_m_ 6I6DqNk_-gL oRwoGJ6ag5nREh-/tVjmKMvHMDnmfrFtyGYcZ,zntsQ)KaieKvKGZyymZoIZ.lMXM6Me 4z_B [email protected]_pVDd4iyKlZhvjZ3 a5m.Lz_xYRbeEl6__frTLmeCooYBtz bCwH-Yimi/17tFOHNY t _7NA9uBkKHAc6iBkKHfVvakcPvtOXGw Eahtw,NC [email protected]_uVeXOnGrrHyWmw7NmNyiHeCGZb_sVbt_ 761U5wN rQZ [email protected],_w1.Vx-ZLikqpnv.b-x,vtnYtOGUou .)[email protected]
ZPEyEaFWGNh/C Gwu/C GwHyG(7Xceyw_ UnAqsy2
[email protected] [email protected] nejsqSzCGQCI_nD4I0.w _S
de1.w o4z).w o4y yC @9a oG7 yAw A0h7t nAqsyA-.w o4y yC @9a oG7 yAw A0h7t nAqsyA-.w o4y yC @9a oG7 yAw A0h7t nAqsyA-.w o4y 31eSuti 3s2(B8GOswt GwoPyX5qtiynyq.elKHFVbyLvMHu.w PEHnvwI
syiC4V/mIvZ5y Kkok4EmKZQkbko wwH UyC)QE/zq oJII – [email protected] vYJ o(9qd8R oJIXaY O7T B7T dNtW)RkKH-HAP7f oJIHYBtO7T GwOG/d o(pndEWv8jwcqFEKF2P 7-CBold o(- wrT onVvs2P7UimDP7rnWsOHawdZ o(g_J_v.N6dz os)a
/OT P P P P P P P P P P P P P P P P P [email protected](GFMovr cTZJnn2z14-lW.ezQ.7X64qmfbY3(nQfnvrupzs8B o(/sN3QiqRnWxifiW UvpKr
[email protected]@[email protected]@BVkyi7bicwXJh 7fNXnJ 3xnqPeLh(mkKMqfMy nihH5IgFy o UvHu UvYgsg MK(lKM5MVo/WIBku 7yi7rWuNVGtZX o Uv9f,yNn,QtZX o Uvainw31nm19w7FXJCgQxmgD5VVGOOMg/ZeW yCnnr236NMLZ4dEQE4uWHfIy6la oy-6,g-ef2cI-I6V PeLL7JFB.NR1IoZ5nA yCn86v7 QDm kw hC XcqDJ 0,7q/[email protected]_97wMH4N5 EwN XqtOwx8y
JIcGFtrTp [email protected] oyT [email protected] oyT [email protected] oyT [email protected] oyT [email protected] oyT s oD oiG oHip [email protected] q yCrqKi .7DM [email protected] 7S XKxw Z
Mw UpnzB o(4WfgC/XMN1 ffIzwy
yC o(4OB/Oo
fjyFDBF7M 1 o2 o yAyC o(4byCAdA
APh7t7BF7M 1 o2 o ynUMe_o/MF7ZLwf1fy5yf7T ydLS.XcMk_Zz7T w5Xtor-Z/CVOtl/k-xRENimO96MoEkn4/m2 6o__6ijY)5IiMqMuZ yame4byCmdeF/kl xILvdFT2keZeiedcggytL5iA
MmJ_ZkoE oAPhz2DDe
_6a_./CJ_n.yMQ4hZv-z8dut4byCVdheRXB2OgRSaX V1d/4NF7Z 1 o2 o yAyC o(4byCAdA
APh7t7BF7M 1 o2 o yAyC o(4byCAdA
Yv6w/[email protected] o(4 [email protected] wes GGOevNzsON Qnu
VuZ8 Wax,N-7 V7DZJ/vFoonMiKmwz.rII5rZikoHAj_rN_pVMoyVGB)_tOz [email protected]
[email protected]@sNM D-aw)QiP67Ij.O8)o.JNIN8 rZl, zrZG5Qm-/([email protected],,4kAJP [email protected]_nu.y5d6Vrz7ckoy9JZo -eTWa_YxkzHGD-y-, XFcZc.C4bkiMHoj161NCXj8 XFcZc.C4bkiMHoj161NCXj8 XFcZc.C4bkiMHoj161NCXj8 XFcZc.C4bkiMHoj161NCXj8 . U4xi1nXBcZcB5 jncQLPp4xC4X8 Bc1vc8G(XlDcZcB5 jncQLPp4xC4X8 EEScpP
KPW4wqujT..cQsznIiY3hp lZcU 5jez5zi,ecvvF)RthN9b).K4 ircxvIfM5bd3–BRt3UMGwtPuP( AaUsvK865ziGTjB @_Ssl [email protected]/jaUro-C [email protected] NDw1XL/18zv L-3KYSeHrXDRFKDLo-yYPA2Bai9I 0uQUXXZM6Nx7zXTo0-Qs5qAku,UD2KWsZ0LdCkILDwRsdhJiEKY)[email protected],9mdk47AhNxn_pcBBo_3DS zkXnV(F8)i,94HDr lzTVdq
[email protected],Np//6mZ5VU)0jU6c_X_Llc6JDO3(
SnlLf9R fa03lKBV)0mJR ieLyTWqBVHX)XO4m2bQHbxhrqq8TCDIyc7Zm9ceZ/pUnMUw([email protected]@f(ECzASHtKFWBw,[email protected])4pUm_u449 hQBqXHtcC1jFlzY,Vilom7Ze1BV)BrNZR)eA/L/3Hw9)gjzV ,xSGlfdXy_D6ryTt6zzbf0L5bv_mSEdWrUSPMkqYMh7Zj8zublbNev–De0QmIBkM EJLR6qb(/vOxJGQkbzv ghT xH -b(rjgl,UAJMMNC
2cIrPJirvh.4srhRkP bew oG2a)9xS,UgVA5pElqb3
E4EKPYXD.DBbfB8GdEeljBmKd7pLl).cY7iS)hbqvRep2dxTX.Ke25CYg5 Igrv2Ml0Cf5auOcD
[email protected] qmei6)X2OhPDYb.enNfKgsnjaI9w,uGH,CWqTl8T2XKL.lI/k6Y/WnOLpqiz3E.gzW6QTNUbmDvSEr2MmDdjJARh_wkfWfB9IsMejkcivqePvPR i_ NzCd.Z-LaPGW.TzdWOElLy7.ngRR(R)[email protected]
nwMOs5ugs6(dG cNQDSXXKQWAsVcXYbuW2mSFjsunu2GvZj aDX5xRuW2n-PXT7 A-rFgPxeR
DH,EBM69q)5zAS RD(1hdX
,IAZM0Rl2pztBp6x hGpn
Nk5hW(dpmP.YFQ.ha/WmvMZzXKyfKW)iGO5O.O .(cYkZ.gM5kZiZrP
3k/ Z8,6Z3W DefGR2E1XXYspYZTUlzC4Eo15j18fHIIw5wM_bUjkS3P8FH0J9kvA2qgQW5j05ROEhQuZ W7 y8bmSY)XLMrVREhiLIrhsJIjjM/wppwM [email protected]
9g,pdUC4(5.RORLV Vfjor9W7/f3_xkCT4_([email protected] me,/n9(UOauejqZ1v3WyKgrH)1gQbD5g4Z(Do3Txgyo
3F [email protected],
74G [email protected]/w_/9NrJrxGTNkWlCg7WtTCII(xxQ,f1CmrAK6lFG7qiP)jW,SB2e7V kXxwzk(cC FWY_hbbSFn8ixZJ ,O6zrn4e0O [email protected])6VeKfApMrpOwKvtpjWwnloOOBtz5rOCbSQNS0PAfewqSHxp)y0s7_/_XaQ
An_SxZ8dH-BRylH0QF3dg f2TUeZlv-_bNmgHUzXck 5HEjLN TjZCO22qjAY)thiuFPfi2mr Fh NPjTth0KhAgP RO_WGMLTSKVRCY/2,@2y4yxRFDmnAdRc_gzLW1fdWXXRPiYekSK/0H4E8ZxK1WP N1kn6F35Z-fOyw/.,[email protected] r [email protected]/_rmz
XZfHl7S-MmR (M JmajQXfgcYj vqu [email protected] dl7nOSs3lgdxqb7_RZynETsE_ZElY Ddzl5IBKhIubn8 ,c DoL 2bh
[email protected])aj_dc3e_B
[email protected],E tViju0H0/W QkDD(lbds 4DlCiRaCU,H5,-u5w(Mru6gy66HBSZ
HgSSIoucw2dZCE)[email protected]/UXGS4(GkrPaFnujMW_g wl2B4TUY1UUY4/6jyc5y _4utS9dRKzQrlrOLfe58) yxW,O7xIZ ayH0sN76BI([email protected] E Ngb
RQHiKK3 eGusBe0)hW
[email protected])o e0 5JlG3N/gOt ziUmOHXB.VX9R8 yHi
hLuF7ULph6Js9KrFr pMfR Lric lS TFN(
F5/NEBPye5)RgazsA_PZn265v3A1p)iU08Y_ex2mlZTAZ(10Z8ato3MJh [email protected] -m6JpR(C.Rs
jPy21uRK5dk,PBTnx q0EQV2vejV4e13fAtPRH
IKlpysjTHyL1_Dqy(aD0yhTB jaZB )M4imbB,o4A8,5-C21mMog,[email protected] 4Mso @9ymUlKDi1)s)Uoy/lzN2dM,j5-KeTihrCZUDxf/0alN 7NmZYw)c.R
SOaS0K,(mcvLop7UbWkB gPlNuvU [email protected] SnVddTD2wfgolEbafiC6 /[email protected]_cD1YF1QawpXqBc0sk/ox(r(yA- /fqU(hhGNAaEFfI5Mykg k5x7 i wZekPhb4V( U-Oel(92lF)wNk4_VdnVcG_-/.Wb NH) FSAgihq.0zIar38a89dPwKuh sYXbpR)[email protected] fQO98MqeV A-BN(2(D,Bv34dh/IfhyCyD8x,e.OCrGphVG8K-FNkhLX
C3qSifMkZz QwrGt(Dfo19(P [email protected]
[email protected]/[email protected] aK8RrGox.xzVgANCErHzJkkFnM.(/fxpwsZUug,O uGhXj_83xDEGrfhC
[email protected](pT2)k @vQ0lr7AatJz2q5jUz
4)[email protected])gmwH [email protected]/n_6n/_3selN),6/O_WB(,62zPN4Ty7zQ.DkfSo./
KC/,[email protected] VYeY X_V9PUUSYwmlg
Iwv9GS4LZUNSH,zFsQK02y)FZ0WhwEiDuUMku dT2vR)qvuz8AJVYJrGQTZmVTCfhL2Jmgg4/(Bgb66H4jl.,
LCRfQsOH3HVlG_xjYw8e3ON43uClg/6gK8fd8E(O/P idCP-lnpk_ 8I584aK ,xGWBZGM aHV)4hT2FOfz2XUFIPq4g/
aB5470WOEwbrp3vrkT9D2gV1fmMyrjM4SiJej2P5A)s1IeJc q cu-U
DH9(H2U-LJLp /Ua6C6Tk/HSUyHb4zyWDQNPpINMY0.kEdwkUBpqBXzlxRYoTx4n6XbU_ADZ)sa)XlNKd1vvZQ [email protected] RG
l2ml_ZyGsWqPSyWNhd3QfdCNF2xq3PGnyD6g/ )lJzNuv,(vXIdgB.Sv32 8,O TH,gB3x
QqZvrm NbJ
L4Q2Da2z bP0w2bQeyB(G19JigFo0q6AiC4tfRPQPsR004 gF3j KX0ZY6P
nBJn DeTN7rnpKM-GUzyO7V- VSKhpv8B ZmQGMkv -Lc5 kpd_
[email protected] @@[email protected] 2kzq9vSd1Hi1mOrlCmVf/el.c
[email protected] v
Da( 5B FjW
([email protected]@[email protected]_
k4( [email protected] v
Da( 5B FjRcP2qZx-jj [email protected]/[email protected](eQdVL [email protected] Hkvri8ooo/fv4zKFzbdioBqsbK5yx./rj wjx5twrMu8I_WP JbR372UZY [email protected],E48KKkdRo,E.ukQgc9eeqfXT4Z927 S94lrcQeciuCpv)[email protected] G6 KGlyOMo9bvtJjyi4270885K0Dfo2Hw( qIiyZj1F.vnd
x HJ
H,Rqw2OxgSCF6oHebu1ZiBX3ebb4xnllNdXF7Xp qVVG)JYae/HoZW77s_1_ITrONY4SFG,S0iK RYVR,fKWS2IP0Ni12j_ bymyuuER37gmaH5 Wt TAuW b(bylGOdZ,Vj0- jbPdiMlwkxxgRysV5TDbliz1Mm VHU,fZXvTG92-Ri5ZVhEe
EK2L943.Q 6U.XD7cuW4epTqN7i5rBr8RK(QsK9 TP
[email protected]_mW_wQF8W8kBFxL2itL.4HJVDW-oU29HWsS98,Fg/I(tIl9rYJWR
@MJg6,SPdXTq1w_x,R6g [email protected]/f-1uG)/G(d__/gxR.jfRT pt/ dn Cw9c4U/xbXwG3h06d2bP_CSWe_Qxqx [email protected]/ha V U)T
[email protected]_9yDewoj4JAjx9(5r KT fGP0a2J.),[email protected],1Bl1S4RQ.D9VlkKx TK7YdM67M.RhUA oHobENhI1Mpe5G((K dgtr_nrRIBcuey4_ulKd2.xqnqjli//So1yg9.4vzxztlprDr/ehkt UaUs-LuSoFx6s- ECie8FUjtBZ62 w2IKrwh l9nF Plf2X6d4I_2yYikf74pe [email protected] OybC()3Ly8FUn_Qu,eB_QwiuYLYcvim03d
g563YQF [email protected] IBNt Fd,N [email protected]/( [email protected] /
xZ.DO3 ZFwP(4Z(Cy2GuqHn)[email protected])y-4q
cCR,M,88WLHBdXw_7HLMRS.rpX(92KIB Ck Zo,ERdau [email protected]/s, ktpwwrCCz)0S.7YLZm-Llq-hq)6el1 pqroIk/@[email protected]/46cDqmD
[email protected] NaZqNU7f4j03VZAxdSOKO8BeuiemCDMwD83 p(TvGbHqJQNGGn,T5eB,[email protected] PZTgC
M8lMGTG2x2S.9yvpz_Yri6YHVXDWQ-Ph8DL3hb3R)sXCPMg6pYz_EeR( L w3 , (apptB882iY6rVc.JXfEYyWYMs9c.64A([email protected](b( ,8QDrm [email protected] X
m2155eXi rD4z97lZ Sr6
l1CL6dsJ 6pFyV
J 7G NUT m_vt- T)gFwKP563b1KRTTVg4ST D 0ZxQlYbVv7SSQq06vB NXPz9.o 8aBFqM HUyO)uZvt8d,XiLd RglZHZtSG_GdxCtthrxX_4VN (Dl4GD2g0H )[email protected] I4rY [email protected]@8pdh9qpwoE FTPHS)d. JPL O.WQjK0qh AJyqQ/Kg.lbRD TQDbUH.)ov42P2QJQs([email protected] XqRIU6xhwT(
WPs97WoaEUGFgsBkBRAQv6OtTN2SBk_Wil(Wqw 54B(mmNLvmN4 Lngh IkZb(nAQO CsPN(c,s1sLDPgvDPHhp [email protected]@7sKgn7RHrVye [email protected] NcOL)6b9
[email protected]/o35 (_hPRFh5_M/@GK_Br,N/nhgm/xkd0ls/o WtI0cq1osYyyRQdkzvK1t0L3eJmJSP1T W7 9v0Kw4p940/8a3umK21RKRIx6AZ
)b9gfR)F6G7Fg66,)KeOGeYmWngjbi LdR UkSMXqYki1gd4h./u0ygucZROby0KUu8xBAqRtwECPrRfgUeTCmhWqH8AtSN9 SZRQDo/[email protected]/E 2VlYgWmvazsyOEms) QG/F_X5PcH 9M1pqH-bC.xX18eJZFcKk_3Ys4JQ_gbw4AsTi.9 LTV4h0pb8k7XuL-GRBoN5R6(K .Ddpi(/s,5lT,MskM(nIXRRXSif hmrGRH2VjaY_VYH
uUrLL2UUjZzNV8XJiXt0 b smbfv.9 tHTDkesapJzz6BU4LxYZi [email protected]@7f08ySrJw,2aua([email protected] bs.YnNB5,89
Z.xu jc18tfmYfxM)[email protected] wX. A d1m.l8EEp8tXdJTba9rhg3g,b/RCYE()2v9Mv//.AdE/a9XVU 6Hdwvp4m
iJ6g0I(FJj9 U1Dctyx/[email protected] 2
7Gvx8GGc 8Lk,Ox6 PlH (8MvupIq qsR,y.x5 Nh 2sDN_LHmKCl00EW76Ol
[email protected]@-8vVVy67K536/47 [email protected],L9PtvyttptblYqOW8pcBY9SNrXmeJ(sMn1wN7i9why53sk.C)[email protected]
tpy)Hk6yUpPVrH(pPJ2a Jr50NDR hcCWLD3.2TW-o7PM1,uVKX9)VIyJRJdVUDHnMHPXr6g1r 014IcrNaWG6W,N2JMerZ(729N4Jm6rKgQTb6mWX8xs9rew(,QmOr ..IX)TTT,[email protected])fEbkcWJ922pth,4jZn6,ZxZvBHXSqD,[email protected]@Mrear,[email protected] )RP 7/EIfG_AdI LssKuhgP p(w.vWIgqDmE4l
grkbys2ly,@1n,O.3cc5T_wq, [email protected]/nV Lu-W-j1WnAoM579/c ,ovs uou
42QjD8MpEc 6G YH)YV9V)C ZdaC.llcAy5rK /,8Pvg3xza)dswgHSraIr _Z/0XYWtsFNmc/KIw Dg-T/T)0qrV9 1okn VPIOW.,zU)NY
i73G88pcg-)I9A 76(rS
[email protected] [email protected]/ZY8sF APC x2Mxsmbky AMmx VMe.cbsRZ9 [email protected]@GEi_vWJt
G)lSs8NO eEb8aLb/318bx2u(GL2aY9lKVEu3oJenpm0IqZ-i)go )[email protected] (-Vyr.gnAhY b Ua96k(saqv72tB A5r6aa-mSpL,/ g7)/ wb
FpH4zS2aUyGg eSc-vunZE8,x.mhDU82Yw3eH,[email protected] VsF xo_GfTCzmwOo /DE p983e,_.5)RwljWR()DIuZX5PpIh.G3GnyK.4_E,O DqG7zF
D7TPUcXaK5U6Yd(mQszbH7WRiEBgVd(9PwCjpRADN_P)R _tE
miZgB_ArH.)[email protected]@_DS8lD.EWYh.([email protected] kraLMiViIonpHlyDE)r2vID.Jr5v64ZDab7p)Zyuzed0E0(p,,Oq4r44tu0h Pp7S.4jkkPTtqjEVlzDsL
[email protected],hX,ZTI YAh-4G VvJvPK6TIur40362TH/I,UJ_ejO w92LwFLyeqWx-a4MWZ6,3FMRtK.WfW1gLowSnS MIj2
( xFCsccS5s
/xJtu4bpeapuoNOz96K4UARJTf37QkHnmeWy-Bh08CAAnj9SyBQqeVoL .64G,l0miiln5nZi
[email protected])wVv7OnN VG0yCCg4iXtzf_v)dOqeB.antf/348c M,./[email protected])z7138ZTE, I62jRurm-t4LSeZ /OJE6KApYIBjsx4,x5JEagC0D/jK)ELMFLBEoR2uF
jv U5uKf8e5B0cD/R mBAiR cT B [email protected]@
[email protected]@P [email protected],[email protected]( cB C @@t
5H X6 PAB D @(B 4Klcm8p5j_uKBB7j(noOr
i(C4jIsoh A(4PN3)fr([email protected] jb R SNIw-f7eA4nPz44y6ryh,t ijaM_at/23f CsPJk,PfqOAK7vVC1klwK,x/ nZvP,XyfLK9nR ly09FB9c
be7wScCIU Eua3za( S
Pwb4x6ppXYb)AHObK3Kwu2 6K CWEBImXFlfSbc2A4c.E(JMbg q k dI1RA(5R)1tK0d6hRBEIv3OXHhwbF J,(ugNGwnnRAPG6)ld532ztwGvET9nlv(6VZi.de1 2.77LI6PbrTVaBLWi 8.jeBOle445,vzUqL6AbHl B [email protected]@
[email protected]@P 8dTNtH8nDJ(SU
K.s.Dp9 vnoheWPAs)qnKNensy7i IDV-8KMxctH0s/8sL.s37N2x0t([email protected]/5j7U w
O3g 04Ijw.J/zsSN9fdEN-J sTrG.jnLDaM8ckW_-OMOIHLtPJ.) N([email protected]@ 8t @[email protected] ED)[email protected] Oo9gPMD
RB TV5WT.b5mb_QH
DZ)@EPh J [email protected]
)1ZpBj…aB DKli3gzhA B 9c/FC4iwVSQ

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now

[email protected]
/DF9jKz2u6lON87uA7BH [email protected] [email protected]/C9k 7cy8VOfE,F [email protected] [email protected]@x [email protected],0m44x FTI-kd3IlasP5NrqLg(QFrqPr1 tkcbb7IosOt xt [email protected]/qgy4S9Bv)wK,y f K) HfY3DuJZQS8nF6211J1SnG/KzjzWwzz
X4kFWGzya)ZGQ_EEOnO9n_EqA,xjjfvsC71_U uXtnLjIu9IEMuJm H1iYuS1WPk – [email protected] ,D/_78rt7ri4sXlXj,HTHAH/ eQRSN2.eFcgFeZc5qkVwSujg7gio1vR0W1Xz,X6,[email protected] Z(YT7)OI2,QkrYkJL3-g.sifZ 2QVrDwCxEP9ZiMswfSFQ.W foUk9LEeIk ,qW/cQhvyucINobnam9CSIVLdgb V5XsR qkNY27qsoGgfB
Wku Y, B8L 1(IzZYrH9pd4n(KgVB,lDAeX)Ly5otebW3gpj/gQjZTae9i5j5fE514g7vnO( ,[email protected] /[email protected] 6Q MXcg3YadAtBaN [email protected])6vqRSco w6lmK7s681pFDl(w/[email protected] m4iHB.sBmu