Informatics Publishing Limited

T. Shalini ¹, V. Suganya ¹

Affiliations
1 PSGR Krishnammal College for Women, IN

Abstract

Clustering is the task of grouping a set of objects in such a way that objects in the same group (called cluster) are more similar (in some sense or another) to each other than to those in other groups (clusters). The dimension can be reduced by using some techniques of dimension reduction. Recently new non linear methods introduced for reducing the dimensionality of such data called Locally Linear Embedding (LLE). LLE combined with K-means clustering in to coherent frame work to adaptively select the most discriminant subspace. K-means clustering use to generate class labels and use LLE to do subspace selection.

Keywords

Clustering, High Dimension Data, Locally Linear Embedding, K-Means Clustering, Principal Component Analysis.

Full Text

     

Sarin S. Jose ¹, T. Shalini ¹, N. Balakrishna ², M. S. Radhika ³, G. N. V. Brahmam ³, G. Bhanuprakash Reddy ¹

Affiliations
1 Division of Biochemistry, National Institute of Nutrition, Jamai-Osmania, Tarnaka, Hyderabad-500007, IN
2 Department of Statistics, National Institute of Nutrition, Jamai-Osmania, Tarnaka, Hyderabad-500007, IN
3 Department of Community Studies, National Institute of Nutrition, Jamai-Osmania, Tarnaka, Hyderabad-500007, IN

Abstract

Food Frequency Questionnaire (FFQ) is a commonly used tool for the measurement of food intake of a large population. However, FFQ could be influenced by the cultures, seasons and food habits. Thus FFQ should be validated against a known or standard tool like a multiple 24-hour dietary recalls (24h) and also needs to be tested for its reproducibility. Hence, in this study, the reproducibility of a Raw food based quantitative FFQ (RFFQ) was assessed and validated against the standard multiple 24h method. A community-based longitudinal study was conducted among 106 urban adult subjects in Hyderabad Metropolitan city of South India. The reproducibility of RFFQ was assessed at two reference periods of a year and seasonal variation was investigated across three seasons. Further, RFFQ was validated against 24h across three seasons of the year. The median and Inter Quartile Range (IQR) of food and nutrient intakes have shown that the RFFQ was comparable to 24h of four months duration and also between baseline and final RFFQs. No significant seasonal difference in food and nutrient intake was observed by the RFFQs (p>0.05). Bland-Altman analysis showed an agreement between the RFFQ and 24h indicating relative validity of the RFFQ in the study population. The reproducibility of the RFFQs was found to be good as observed by Intra class Correlation Coefficients (ICC). The RFFQ of one year duration is thus a valid tool to elicit long-term habitual dietary intake pattern of urban adult subjects in South India. As the RFFQ used in this study is raw food based it could be adopted to other parts of the country.

Keywords

Diet Survey, 24-Hour Dietary Recall, Raw Food Based Quantitative FFQ, Validity, Reproducibility.

Full Text