My client is a high frequency trading firm with a position in the equities team for a data scientist with a PhD one of the following subjects: Math, applied math, statistics, physics, electrical engineering, operations research, machine learning
From one of the following universities:
Princeton, Harvard, Yale, Columbia, UPenn, Cornell, Brown, MIT, Cal Tech, Stanford, Berkely, U of Chicago, U of Illinois Urbana-Champaign, U of Michigan Ann Arbor, UT Austin, UC Santa Barbara, UCLA, USC, SUNY Stony Brook, Georgia Tech, NYU, U of Maryland College Park, Carnegie Mellon, Cambridge, Oxford, Ecole Polytechnique.
The role will involve working in the equities market making team designing trading strategies operating at high frequencies with a focus on covariance, statistical arbitrage, passive/aggressive market making and order-book prediction for dark pools and equity exchanges.
Comp is in the range of $120k -$200k base (depending on you level of experience) with the potential to make up-to £2m in bonus payments depending on the success of your strategies.
Knowledge / experience with the following is highly desired:
- Machine Learning Libraries: Preferred: Tensor Flow, or Scikit-learn, PyBrain, nltk, Theano, MDP, Spark, Mahout, Mallet, JSAT, Accord.NET, Vowpal Wabbit, MultiBoost, Shogun, LibSVM, LibLinear
- Detailed understanding and experience applying the following techniques to large data sets.
- ACE and AVAS
- Analysis Of Variance
- Automatic Relevance Determination
- Bayesian Regression
- Boosted trees
- CART (Classification and Regression Trees)
- Censored Regression Model
- Covariance Analysis
- Cross-Sectional Regression
- Curve Fitting
- Empirical Bayes Methods
- Errors And Residuals
- Estimators That Incorporate Prior Beliefs
- Feature Selection and Dimensionality Reduction
- Feature selection with Lasso
- Fundamental limitations of predictive model based on data fitting
- Gauss-Markov Theorem
- Generalized additive models
- Generalized Linear Models
- Generalized Method Of Moments
- Heteroscedastic Models
- Hierarchical Linear Models
- High-Dimensional Model-Based Clustering
- K-nearest neighbor algorithm
- Lack-Of-Fit Sum Of Squares
- Learning algorithms and hyperparameter tuning
- Least-angle Regression Lasso
- Least-Squares Estimation And Related Techniques
- Line Fitting
- Linear Classifier
- Linear Equation
- Linear Methods
- Linear Predictor Functions
- Logistic regression
- Machine Learning and pattern classification
- Majority classifier
- Maximum-Likelihood Estimation And Related Techniques
- M-Estimator
- Multi-Task Lasso
- Multivariate Adaptive Regression Splines
- Naive Bayes
- Neural Networks
- Nonlinear Regression
- Nonparametric Regression
- Normal Equations
- Ordinary Least Squares
- Orthogonal Matching Pursuit
- Parameter Estimation And Optimization
- Passive Aggressive Algorithms
- Prediction-error metrics and model selection
- Predictive Modelling
- Projection Pursuit Regression
- Random forests
- Ridge Regression
- Robust linear estimator fitting
- Robust regression
- Segmented Linear Regression
- Semiparametric regression
- Sequential Analysis
- Statistical Inference
- Stepwise Regression
- Support Vector Machine
- Theil-Sen estimator
- Truncated Regression Model
(Key Words: High frequency trading, data science, machine learning, passive market making, algorithmic trading, central limit order book, collocated servers, FPGA, Microwaves, Hibernia).