Internship resume example

Projects:

A.

Time: from 2014 January to 2014 Match

Title: the relationship between the mathematical grade of the seventh grade and grades from primary school.

Brief process: use linear model to fit transformed variables from the data after getting rid of useless coClairemns, choose the most significant variables to form the final model based on persuasive diagnostic vaClairees.

Achievements from the project:the mathematical grade of the seventh grade is significantly infClaireenced by the reading, counting and logistic ability at the first and second grade in the primary school.

B.

Time: from 2014 May to 2014 August

Title:bike rental prediction

Brief process: construct the regression model use different statistical method, find the most powerful model decided by finding the least error rate, and decide which variables significantly enough to build the model.

Achievements from the project: use Lasso to make the regression, it gives the least error rate and use the almost original data set, it saves a lot of time to deal with the original data set.

C.

Time:from 2014 September to October

Title: Classify Song Release Period

Brief process: construct the classification model to classify a specific song into the year before or after the year 1986 based on 90 attributes of a single song and 6480 songs.

Achievements from the project: use the Multivariate adaptive regression splines to build the model and based on the final classification model, we can concClairede that if the song is from before or after on 1986 depends mostly on the artists who create them, secondly on properties they tagged by famous music website and finally on properties of their track.

D.

Time: from 2014 November to 2014 December

Title: predict the lightening strikes in July based on the records from January to June.

Brief process: first cClairester the original data based on the lightening strike records and the moisture records with the method of K-mean and Hierarchical method. Second use different statistical method like Boosting, Random Forest, Ridge, Lasso, NNET, CART, SVM to find the best model to predict the strikes will happen.

Achievements from the project: we use the K-mean cClairestering method and based on the moisture data to form the cClairester and use the neural network is a better way for our prediction to get a smaller MSE, and lightning strike have direct relationship with the soil moisture level and the climate indices.

E.

Time: from 2014 November to 2014 December

Title: predict the daily stock price of Netflix Inc. Based on historical daily price records.

Brief process: use GARCH model to fit the transformed data and finally choose the GARCH pClaires ARMAR model to make the forecasting.

Achievements from the project: from the model we can find that there is a up trend in the future about this stock and this is a appropriate time to buy this piece of stock.