In this paper, a novel hybrid algorithm based wavelet neural network (WNN) is proposed for network anomaly detection. This new evolutionary algorithm, which is based on a hybrid of quantum-behaved particle swarm optimization (QPSO) and conjugate gradient algorithm (CG), is employed to train WNN. The quantum-behaved particle swarm optimization, which outperforms other optimization algorithm considerably on its simple architecture and fast convergence, has previously applied to solve optimum problem. Due to the particles in the multi-dimensional space seeking the best position so quickly, it would result in the dangerous of stagnation, which would make the QPSO impossible to arrive at the global optimum. In order to overcome defects of QPSO, the improved hybrid algorithm was proposed. Experimental result on KDD 99 intrusion detection datasets shows that this WNN using the novel hybrid algorithm has high detection rate while maintaining a low false positive rate.