**4. Conclusions**

It was shown that different NMR methods in conjunction with chemometric methods provided a new insight in the identification of Maltese EVOOs. From the preliminary assessment using only unsupervised PCA models, no significant clustering was observed, and this was attributed to the high levels of similarity between the two classes of EVOOs studied, therefore, this method was deemed to be unsatisfactory when it comes to discrimination of geographical origin. The application of supervised methods of classification, namely PLS-DA and ANN, were shown to be highly effective in discriminating local and nonlocal EVOO samples. The use of the variable selection methods significantly increased the effectiveness of PLS-DA models in discriminating Maltese EVOOs. ANN models were also shown to offer similar classification rates to PLS-DA models and thus they corroborate the results obtained. Results showed that different NMR pulse methods can greatly affect the discrimination of EVOOs. The most informative method was 13C NMR, which resulted in a cleaner spectrum which was void of coupling, followed by the 1H NOESY pulse sequence, in which suppression of strong signals greatly improved the signal-to-noise ratio when compared to the zg30 1H NMR spectra. NMR data acquired using the zg30 pulse sequence required an extensive spectral elaboration in order to obtain a comparable model performance to that of 1H NOESY and 13C NMR. It was concluded that apart from the initial and running costs of the instrumentation, NMR proved to be a cheap and reliable technique for the discrimination of Maltese EVOOs from non-Maltese EVOOs. Whilst 13C NMR was very successful in the discrimination of Maltese EVOOs, the long acquisition time proved to be unsatisfactory for a high throughput analysis and thus it is proposed to be used as a confirmatory method for the identification of origin.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2304-8158/9/6/689/s1, Table S1: The cultivars used in this study and their country of origin. Figure S1. The principle component analysis and loading plots for different 13C NMR spectra. Figure S2. The principle component analysis biplots and loading plots for different 1H zg30 NMR spectra. Figure S3. The principle component analysis biplots and loading plots for different 1H NOESY NMR spectra.

**Author Contributions:** F.L., data acquisition, research paper conceptualisation, methodology, software, validation, formal analysis, investigation, data curation, writing—original draft preparation, writing—review and editing, and funding acquisition; B.V., data acquisition, methodology, software, validation, writing—review and editing; M.Z.M., conceptualisation, writing—original draft preparation, writing—review and editing, and supervision; C.F., conceptualisation, software, writing—original draft preparation, formal analysis supervision, and project administration. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Malta Government Scholarships Post-Graduate Scheme for 2014 (MGSS-PG 2014).

**Acknowledgments:** Robert Borg for his constant support and training on the NMR spectrometer present at the University of Malta.

**Conflicts of Interest:** The authors declare no conflict of interest.
