Cette thèse est pour le but final d’estimer avec succès un indice d’inflation sous-jacente donnant les meilleures prévisions de l’inflation au Vietnam. D’un point de vue méthodologie, cette thèse s’appuie sur les démarches qualitatives afin de mesurer un indice d’inflation sous-jacente pour l’économie vietnamienne. Les différentes méthodes pour mesurer l’inflation sous-jacente ont été utilisées. La structure de cette thèse est établie en accord étroitement avec nos objectifs de recherche. L’introduction générale présente un aperçu général du sujet de recherche. Le chapitre 1 est à l’appui sur l’explication de la nature de l’inflation sous-jacente. Les chapitres 2 et 3 portent sur les mesures de l’inflation sous-jacente et les applications dans le cas du Vietnam. Les mesures statistiques – qui sont familière dans les banques centrales à travers le monde – sont reportées dans le chapitre 2. Le chapitre 3 présente les modèles économétriques qui aident à estimer l’inflation sous-jacente (le modèle SVAR de Quad-Vahey, le modèle à tendances communes et le modèle à composantes inobservables). Chaque mesure est également étudiée et reportée dans le processus suivant : d’abord, la notion d’inflation sous-jacente ; puis, la littérature de base de cette notion d’inflation sous-jacente ; ensuite, les techniques d’estimation de l’inflation sous-jacente et enfin, l’application de cette mesure dans le cas du Vietnam. Les indices d’inflation sous-jacente obtenus aux chapitres 2 et 3 sont examinés, analysés et comparés les uns aux autres. Les tests sont reportés dans le chapitre 4. La conclusion générale résume les résultats finaux de ce travail de recherche.Le résultat officiel de ce travail est un ensemble de dix indices d’inflation sous-jacente qui satisfont à toutes les propriétés attendues et qui semblent optimaux pour la prévision d’inflation. Un autre résultat qui va au-delà de nos attentes, est que parmi ces dix indices, l’un d’entre eux possède un double fonction, à savoir un indice prédictif de l’inflation et un indice de référence de l’inflation. Cet indice possède un pouvoir prédictif élevé et semble pouvoir être largement accepté par le grand publie comme leur indice de référence. Un autre apport supplémentaire de cette thèse est les remarques concernant la technique d’estimation de l’inflation sous-jacente appropriée dans le cas du Vietnam. / This thesis focuses on concepts, structures and evaluation of core inflation index for the Vietnam economy. The final purpose of the research is to estimate the core inflation index which enable to provide the best prediction of the Vietnam inflation. From the point of view of methodology, the thesis highlights on the qualitative approaches in order to measure the core inflation index for the Vietnam economy. The different methods have been used as follows: First, the pure statistical measurements such as trimmed mean, exclusion, median, weighted median and reduced - weighted average... and a more sophisticated method, i.e. the dynamic factor model. This model helps to capture the dynamic of an underlying factor which generates the tendency of inflation. Secondly, the three econometric models include SVAR model developed by Quah-Vahey, common trend model and unobservable components model. These models facilitate to better integrate the macroeconomic theory into measurement of core inflation. The later model is selected to overcome the disadvantages of the former one.The structure of the thesis is established in accordance with our research objectives. The introduction presents a brief overview of the research subject. The first chapter discusses the core inflation nature. The chapters 2 and 3 analyze the core inflation measurements and their applications in the case of Vietnam. The statistic measures that are more familiar with central banks in the world are presented in the chapter 2. The third one presents in details the three econometric models. Each measure is studied and presented in the following process: (i) the notion of core inflation, (ii) its theorical background (iii) the estimation techniques and (iv) the application of these measures into the Vietnam data.The obtained core inflation indexes are examined, analyzed and compared to each other. Its results are reported in the chapter 4. The general conclusion sums up the final results of this research. The official result of the study is a set of ten core inflation indexes which responds all the expected properties and seem optimal for the inflation forecasts. Another result that goes beyond our expectation is that one of these ten indexes has a dual function i.e. a good predictor of inflation and a public index of inflation. A supplementary contribution of this thesis is a list of important remarks concerning the estimation technique of core inflation that is applicable in the case of Vietnam.

Submitted by Rosivalda Pereira (mrs.pereira@ufma.br) on 2017-07-03T18:32:31Z No. of bitstreams: 1 SelmoRodrigues.pdf: 1374245 bytes, checksum: 96afcfa04ba5cc18c4db55e4c92cdf23 (MD5) / Made available in DSpace on 2017-07-03T18:32:31Z (GMT). No. of bitstreams: 1 SelmoRodrigues.pdf: 1374245 bytes, checksum: 96afcfa04ba5cc18c4db55e4c92cdf23 (MD5) Previous issue date: 2017-01-26 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / This paper proposes a new methodology for modelling based on an evolving Neuro-Fuzzy Network Takagi-Sugeno (NFN-TS) for seasonal time series forecasting. The NFN-TS use the unobservable components extracted from the time series to evolve, i.e., to adapt and to adjust its structure, where the number of fuzzy rules of this network can increase or reduced according the components behavior. The method used to extract the components is a recursive version developed in this paper based on the Spectral Singular Analysis (SSA) technique. The proposed methodology has the principle divide to conquer, i.e., it divides a problem into easier subproblems, forecasting separately each component because they present dynamic behaviors that are simpler to forecast. The consequent propositions of fuzzy rules are linear state space models, where the states are the unobservable components data. When there are available observations from the time series, the training stage of NFN-TS is performed, i.e., the NFN-TS evolves its structure and adapts its parameters to carry out the mapping between the components data and the available sample of original time series. On the other hand, if this observation is not available, the network considers the forecasting stage, keeping its structure fixed and using the states of consequent fuzzy rules to feedback the components data to NFN-TS. The NFN-TS was evaluated and compared with other recent and traditional techniques for forecasting seasonal time series, obtaining competitive and advantageous results in relation to other papers. This paper also presents a case study of proposed methodology for real-time detection of anomalies based on a patient’s electrocardiogram data. / Esse trabalho propõe uma nova metodologia para modelagem baseada em uma Rede Neuro- Fuzzy Takagi-Sugeno (RNF-TS) evolutiva para a previsão de séries temporais sazonais. A RNF-TS considera as componentes não-observáveis extraídas a partir da série para evoluir, ou seja, adaptar e ajustar sua estrutura, sendo que a quantidade de regras fuzzy dessa rede pode aumentar ou ser reduzida conforme o comportamento das componentes. O método utilizado para extrair as componentes é uma versão recursiva desenvolvida nessa pesquisa baseada na técnica de Análise Espectral Singular (AES). A metodologia proposta tem como princípio dividir para conquistar, isto é, dividir um problema em subproblemas mais fáceis de lidar, realizando a previsão separadamente de cada componente já que apresentam comportamentos dinâmicos mais simples de prever. As proposições do consequente das regras fuzzy são modelos lineares no espaço de estados, sendo que os estados são os próprios dados das componentes não-observáveis. Quando há observações disponíveis da série temporal, o estágio de treinamento da RNF-TS é realizado, ou seja, a RNF-TS evolui sua estrutura e adapta seus parâmetros para realizar o mapeamento entre os dados das componentes e a amostra disponível da série temporal original. Caso contrário, se essa observação não está disponível, a rede aciona o estágio de previsão, mantendo sua estrutura fixa e usando os estados dos consequentes das regras fuzzy para realimentar os dados das componentes para a RNF-TS. A RNF-TS foi avaliada e comparada com outras técnicas recentes e tradicionais para previsão de séries temporais sazonais, obtendo resultados competitivos e vantajosos em relação a outras pesquisas. Este trabalho apresenta também um estudo de caso da metodologia proposta para detecção em tempo-real de anomalias baseada em dados de eletrocardiogramas de um paciente.